What's hot in molecular diagnostics and biomarkers? AACR #1
One of the biggest challenges with Vascular Endothelial Growth Factor (VEGF) therapy to date has been the singular lack of either predictive or prognostic biomarkers.
This means that we have no idea which patients are most likely to respond to therapy (ie predictive) when selecting either a monoclonal antibody (eg bevacizumab) or a small molecule tyrosine kinase (eg sorafenib, sunitinib or pazopanib), nor do we will know what their likely prognostic outcome might be in terms of survival.
In an ideal world, we would be able to predetermine and monitor therapy for specific subtypes, thereby avoiding exposing thousands of patients to the systemic effects (and costs) of a drug that may not work for them.
Of course, we all know that developing biomarkers is:
- not easy
- could be rather expensive
I was therefore greatly cheered while at the AACR meeting Denver on Molecular Diagnostics and Cancer Therapeutics last week to come across a little gem of a poster from the scientists at AVEO Pharmaceuticals.
AVEO ($AVEO), a biotech based in Boston, are developing an oral VEGF inhibitor called tivozanib, currently in phase III for renal cell cancer, and appear to have developed a method to predict which patients are more likely to respond to the compound. Whoa!
I've been watching this company for a couple of years now and have been impressed with what I see so far. Two years ago I met their CMO Bill Slichenmyer over lunch at the AACR meeting in Denver when he was at Merrimack Pharma and kept track of what was happening at AVEO when he moved there. Both companies have interesting technology platforms and smart scientists.
Aside from the poster, AVEO's head of translational medicine, Murray Robinson, also presented the data during an oral session. What was particularly interesting was that the findings were not what one might expect – at all. They wondered if the potential biomarker they identified in animal studies might be reproduced in humans.
AVEO found their biomarker by inserting specific oncogenes and other engineered genes altered in numerous cancer types into the tissue of animals then studying the variety of tumours that were produced. One example of this approach was to genetically alter the HER2 gene, resulting in tumours that naturally expressed different pathways for growth.
They then looked at 600 tumour samples in clinical trials across eight different tumour types and realised that essentially the same biomarker identified in their breast tumour model was indeed associated with clinical activity in a set of kidney tumour patients from a previous Phase II kidney cancer trial. This biomarker was associated with white blood immune cells that are recruited into the tumour to produce angiogenic growth factors and leads to intrinsic resistance to tivozanib.
I confess to being kind of awed by this sort of research.
For some time, clinicians have been grumbling about not having a biomarker for Avastin, Sutent or Nexavar to better help choose which patients would be most likely to respond, thereby avoiding the need to treat everyone to gain a benefit in a few. Here we have three big pharma companies and no biomarker. A little biotech comes along with some smart ideas, a rational approach to the problem and some creative thinking to developing a biomarker for their compound, which is not yet on the market.
Of course, this biomarker is specific to AVEO's tivozanib, as no work has been completed to show that the myeloid component they identified is relevant in the others.
The good thing is that it's now the first biomarker associated with a VEGF therapy.
The bad news is that we will have to wait a little longer to see if the results of the phase III trials in kidney cancer are good enough for approval, but hopefully that won't be too long now.
Imagine one step further.
Currently, the FDA is reviewing Roche's Avastin in breast cancer and deciding whether or not to withdraw the application given the marginal data currently available from trials such as AVADO. Suppose Roche/Genentech had a biomarker that was relative to Avastin and could be helpful for either prognostic or better still, predictive purposes? Then you could actually make better use of the drug based on a biomarker.
Before anyone in big pharma jumps up and down and starts moaning about the cost and the difficulty, take a look at AVEO's logical, sensible technical approach to the problem. You realise that what we really need is more imagination and creativity in R&D and less objections to progress.
Now suppose the biomarker AVEO identified in their breast cancer models turns out to be useful in breast cancer for women on their compound? If you can clearly show an association between different subsets, who is likely to develop resistance and who is more likely to respond, plus better outcomes, what's not to like? The overall response rates will be higher in some subsets and lower in others, rather than a crapshoot of "well, it helps some women" or how about the vague "many women clearly benefitted". Great, but which ones and why?
In my opinion, AVEO have done a great job identifying a relevant biomarker for their compound which may actually increase rather than lessen the chances of successful approval down the road.
May the force be with them!
5 Responses to “What's hot in molecular diagnostics and biomarkers? AACR #1”
Why hasn’t there been any progress at all in drug selection through the use of molecular diagnostics and biomarkers? Simply put, they do not work! Little progress has been made in identifying which therapeutic strategies are likely to be effective for individual patients by molecular prognostic and predictive markers.
It was hoped that any patient with cancer would have their tumor biopsied and profiled. The profile would then be displayed as a unique genetic signature, which would in turn predict which therapy would most likely work. However, gene-expression signatures are not ready for prime time.
Although molecular profiling of tumors has led to the identification of gene-expression patterns, a new review published in the March 16, 2010 JNCI has found little evidence that any of the signatures are ready for use in the clinical setting.
And then on July 22 and 23, 2010, the NCI suspended three clinical trials designed to apply gene profiling to select treatments for patients. The first chink in the armor came when scientific reviewers issued an “expression of concern” regarding the validity of the entire method.
Further analyses revealed evidence that the technologies for the prediction of response in individual patients could not be reproduced. The NCI convened a group of 31 scientists, who concluded it’s absolutely premature to use these prediction models to influence the therapeutic options open to cancer patients. The genomic methodology is not ready for clinical application.
What went wrong?
“The simple answer is that cancer isn’t simple,” according to Dr. Robert Nagourney, one of the pioneers of functional profiling analysis.
Cancer dynamics are not linear. Cancer biology does not conform to the dictates of molecular biologists. Once again, we are forced to confront the realization that genotype does not equal phenotype.
In a nutshell, cancer cells utilize cross talk and redundancy to circumvent therapies. They back up, zig-zag and move in reverse, regardless of what the sign posts say. Using genomic signatures to predict response is like saying that Dr. Seuss and Shakespeare are truly the same because they use the same words. The building blocks of human biology are carefully construed into the complexities that we recognize as human beings.
However appealing gene profiling may appear to those engaged in this field (such as Response Genetics, Caris, the group from Duke and many others) it will be years, perhaps decades, before these profiles can approximate the vagaries of human cancer.
Functional profiling analyses, which measure biological signals rather than DNA indicators, will continue to provide clinically validated information and play an important role in cancer drug selection. The data that support functional profiling analyses is demonstrably greater and more compelling than any data currently generated from DNA analyses.
Functional profiling remains the most validated technique for selecting effective therapies for cancer patients. There is a microvascular viability assay with functional profiling for anti-angiogenesis-related drugs (Avastin, Sutent and Nexavar), developed by another pioneer of functional profiling analysis, Dr. Larry Weisenthal.
A major modification of the DISC (cell death) assay allows for the study of anti-microvascular drug effects of standard and targeted agents, such as Avastin, Nexavar and Sutent. The microvascular viability assay is based upon the principle that microvascular (endothelial and associated) cells are present in tumor cell microclusters obtained from solid tumor specimens. The assay which has a morphological endpoint, allows for visualization of both tumor and microvascular cells and direct assessment of both anti-tumor and anti-microvascular drug effect. CD31 cytoplasmic staining confirms morphological identification of microcapillary cells in a tumor microcluster.
The principles and methods used in the microvascular viability assay include: 1. Obtaining a tissue, blood, bone marrow or malignant fluid specimen from an individual cancer patient. 2. Exposing viable tumor cells to anti-neoplastic drugs. 3. Measuring absolute in vitro drug effect. 4. Finding a statistical comparision of in vitro drug effect to an index standard, yielding an individualized pattern of relative drug activity. 5. Information obtained is used to aid in selecting from among otherwise qualified candidate drugs.
It is the only assay which involves direct visualization of the cancer cells at endpoint, allowing for accurate assessment of drug activity, discriminating tumor from non-tumor cells, and providing a permanent archival record, which improves quality, serves as control, and assesses dose response in vitro.
Photomicrographs of the assay can show that some clones of tumor cells don’t accumulate the drug. These cells won’t get killed by it. The assay measures the net effect of everything which goes on (Functional Tumor Cell Profiling methodology). Are the cells ultimately killed, or aren’t they?
This kind of technique exists today and might be very valuable, especially when active chemoagents are limited in a particular disease, giving more credence to testing the tumor first. After all, cutting-edge techniques can often provide superior results over tried-and true methods that have been around for many years.
Bibliography relevant to AngioRx/Microvascular Viability assay (MVVA)
1. Weisenthal, L. M. Patel,N., Rueff-Weisenthal, C. (2008). “Cell culture detection of microvascular cell death in clinical specimens of human neoplasms and peripheral blood.” J Intern Med 264(3): 275-287.
2. Weisenthal, L., Lee,DJ, and Patel,N. (2008). Antivascular activity of lapatinib and bevacizumab in primary microcluster cultures of breast cancer and other human neoplasms. ASCO 2008 Breast Cancer Symposium. Washington, D.C.: Abstract # 166. Slide presentation at: http://tinyurl.com/weisenthal-breast-lapatinib
3. Weisenthal, L. M. (2010). Antitumor and anti-microvascular effects of sorafenib in fresh human tumor culture in comparison with other putative tyrosine kinase inhibitors. J Clin Oncol 28, 2010 (suppl; abstr e13617)
4. Weisenthal, L., H. Liu, Rueff-Weisenthal, C. (2010). “Death of human tumor endothelial cells in vitro through a probable calcium-associated mechanism induced by bevacizumab and detected via a novel method.” Nature Precedings 28 May 2010. from http://precedings.nature.com/documents/4499/version/1
Actually, in simple terms Gregory, I disagree with your fundamental premise. That said, I’m not going to get into a war of words here, but I would appreciate any future comments you make be short and to the point, otherwise why not go and set up your own blog and have a voice there?
Many thanks.
[…] as a useful biomarker of response for tivozanib in renal cell cancer. You can read more about that here if you missed it. # This raises several interesting questions for me: […]
There are selected groups who will benefit from Avastin, if they knew who they were. Roche researchers have been looking for tests to help predict how patients will respond to Avastin. Some have suggested that they should use the cell-based functional profiling platform (AngioRx Assay) to identify a potential targeted population of cancer patients that it thinks will benefit from Avastin, and then conduct a randomized clinical trial among this group.However, unlike some genetic assays that look whether an individual has a particular mutation or amplification, and therefore tests for “theoretical” candidates for a particular targeted drug, the functional profiling technique may find Avastin not synergistic and finds some other VEGF-targeted (or multiple VEGF-targeted) drug may work better in an individual cancer patient and then put that individual into the clinical trial. I can understand they may not want some other drug tested on their dime.
This reinforces the need for cancer therapies to be individualized. It remines us that it is the good outcome of the patient not the therapy applied that constitute successful therapy. There is really nothing wrong with Avastin. It’s a wonderful drug that incorporates the brilliant insights originally articulated by Judah Folkman. There are not perfect drugs. There are simply drugs that work for certain patients. But that’s not what pharmaceutical companys like to hear. They like to produce drugs that apply to a broad base of patients. To make the most out of a drug, not just some subsets of patients.
[…] thing to consider is that previously, AVEO presented some rather promising data on biomarkers of response in their preclinical and phase I RCC data. I’m very keen to see if any subset analysis in the […]
Comments are closed.