Finally a Way to Predict Bevacizumab Response?
 

David Kerr, Professor of Cancer Medicine at University of Oxford, United Kingdom, and past President of ESMO, wrote about an article published in the Journal of Clinical Oncology by Lambrechts and colleagues, which looked at predictive markers for the antiangiogenic drug bevacizumab.

All of us understand the impact that bevacizumab has had across a range of metastatic tumor types on slowing the progression of disease. This drug is widely used around the world. One of the "holy grails" has involved finding markers that would allow us to select and identify patients who will benefit most. A huge amount of work has been performed to find such markers, and in this article the author has managed to distill the research to give us an idea about which of these many markers may be most successful in the clinic. The candidates actually turn out to be quite a small number of markers.

When we consider circulating markers, then short-form VEGF-A (vascular endothelial growth factor-A) looks as though it may be a likely candidate. When we look at the tumor tissue itself, expression of VEGFR-1 -- the vascular endothelial growth factor type 1 receptor -- if it is expressed at low levels, is a marker of greater likelihood of response and clinical benefit from treatment with bevacizumab. Similarly, tumor expression of low levels of neuropilin-1 has the same effect.

Some genetic work looking at the germline response is consistent with this. A set of SNPs, single nucleotide polymorphisms, in the genes encodes VEGFR-1. Those SNPs, which are associated with relatively high levels of expression of the receptor, are associated with a lower chance of response to bevacizumab. Thus, we see an internal consistency. Not only are these the markers that have been statistically identified as being those most likely to help us select patients who will benefit most, but there is also an entirely plausible underlying biology that stretches all the way from preclinical studies through to the clinic that supports use of these markers.

At the moment, these markers are not validated because no prospective study has stratified or selected patients on the basis of these markers. These authors propose to undertake a study, which they call MERiDiAN, in patients with metastatic breast cancer in which they will stratify patients on the basis of circulating levels of short VEGF-A. This will be the proof of the pudding and may at last shed some interesting light on what has proven an elusive concept: the ability to identify markers for a drug like bevacizumab, which functions at the interface between the tumor and the stroma and therefore affects those cells that are most easy to manipulate -- cancer cells, whether cell lines, explants, and so on. The answer does not lie entirely there, hence the relative difficulty.

Nevertheless, this is a masterful review. The authors collected the data beautifully and, best of all, tell us what we need to do next in order to be able to validate these findings and take a further step toward clinically validating a useful marker for this important drug.

References

Lambrechts D, Lenz HJ, de Haas S, et al. Markers of response for the antiangiogenic agent bevacizumab. J Clin Oncol. 2013;31:1219-1230.

Biomarkers of Antiangiogenic Therapy: How Do We Move From Candidate Biomarkers to Valid Biomarkers?

http://cancerfocus.org/forum/showthread.php?t=3372

French interpretation of this
 

Bevacizumab has its own activity, usually in combination with cytotoxic in many neoplasia (breast, kidney, ovarian, colorectal cancer).

The impact is most often measured by a small but significant gain in DFS with heterogeneity in cancer and between populations.

However, the lack of biological factors identified for the response, a non-anecdotal toxicity profile, and the high cost of the product, led U.S. regulators to revoke its use in cancer metastatic breast, pending further data, including biological type.

Bevacizumab is not, strictly speaking, a targeted therapy, since its action and target are relatively ubiquitous. However, it should be possible to identify biomarkers of efficacy. Thus the authors of this beautiful literature review suggest all tracks being followed.

The main target of bevacizumab is the theory circulating VEGF-A. Many other factors involved in angiogenesis and stimulation of pericytes potentially involved in the mechanism (IGF, FGF, etc. ..).

Regarding circulating factors, it identifies 17 randomized studies have studied how the correlation between the ancillary level of expression of certain biomarkers with efficacy parameters (DFS or OS). It follows from this that the short isoform of VEGF-A is the candidate most "sharp" to predict the DFS expression of neuropilin-1, and the expression of VEGF receptor-1 in plasma and tumor are also good candidates.

Similarly, several studies have identified changes in the rate of VEGF-A and PlGF as a predictor of response.

In the tumor environment, the study of VEGR1 markers and NRP1 was not real contribution, but remains an interesting way.

Genetic variants of VEGF-A are also avenues to explore. Indeed, more than a dozen variants have been identified as correlated with PFS in twenty studies.

Regarding tumor subtypes in breast cancer, studies AVADO, BEATRICE and AVEREL failed to 'actually identify a correlation.

In colon cancer, genomic status of KRAS and BRAF does not appear to be related to efficiency.

A first study has been ongoing since 2012 in metastatic breast cancer, named Meridian with stratification based on the expression level of the short isoform of VEGF-A. This is really the first study based its main hypothesis on the relationship between biomarker and efficacy.

I bet that a number of studies done on this concept will be created, which is not necessarily obvious in the priorities of development time and the high cost of this type of study on a scale Phase III.

David Coeffic, Editor France Cancer

Reference: Lambrechts D, Lenz HJ, Haas S, et al. Markers of response for the antiangiogenic agent bevacizumab. J Clin Oncol. 20 March 2013, 31 (9) :1219-30.