CSN Login
Members Online: 10

You are here

GOG Meeting

gdpawel's picture
Posts: 538
Joined: May 2001

The Gynecologic Oncology Group (GOG) held its 84th Semi-Annual Meeting on Januray 26-29, 2012 in San Diego, CA.


Presented last summer, in the first head-to-head clinical trial comparing gene expression patterns (molecular profiling) with personalized cancer cytometric testing (functional profiling or chemosensitivity testing), personalized cancer cytometrics was found to be substantially more accurate.

Four different genes were included in the molecular part of the study. The four genes were selected as those which researchers believe to have the greatest likelihood of accurately predicting individual patient response to specific anti-cancer drugs.

Although this was the first head-to-head, the accuracy levels found in this trial for personalized cancer cytometric testing were strikingly consistent with those documented in dozens of previous studies, published by respected cancer researchers around the world. In those studies, as in this one, extremely high levels of correlation (in other words, high levels of test accuracy) were found for personalized cancer cytometrics.

Source: Arienti et al. Peritoneal carcinomatosis from ovarian cancer: chemosensitivity test and tissue markers as predictors of response to chemotherapy. Journal of Translational Medicine 2011, 9:94.

It was hoped that something like this study would be proposed at the semi-annual GOG meeting. Perhaps a good three-armed clinical trial: physician's choice vs molecular profiling vs functional profiling?

I understand this issue has been an ongoing saga for some years now. The question that should be addressed is an across the board assessment of the relative accuracy of different endpoints. They should answer all the questions at once, not one followed by another. I hope we do not have to wait until the summer ASCO meeting to find out.


carolenk's picture
Posts: 909
Joined: Feb 2011

Thanks, Greg, for posting this. I hope we see further progress--soon.

gdpawel's picture
Posts: 538
Joined: May 2001

On ASCO's Connection, a professional networking site for ASCO's worldwide oncology community, Don S. Dizon, MD, FACP posted his take on the semi-annual GOG meeting.

He was engaged in updates of ongoing clinical trials and discussion of new trial concepts for ovarian, cervix, endometrial, and other gynecologic cancers. He learned that we are entering a new era in clinical trials—gone are the days of doing a trial in gynecologic cancer using the "next best drug" because it worked in (whatever) cancer. No—the onus on investigators is to design new trials based on science. As Dizon's mentor and friend, Dr. Carol Aghajanian (chair of the Developmental Therapeutics Committee) put it: "No tissue, no marker, no target—no trial."

During the meeting, he also found himself considering the endpoints of clinical trials: should it be the response rate? Should it be the proportion of patients who have no tumor growth at six months? Or should it be overall survival? What about quality of life—could this ever be a primary endpoint in clinical trials? What about toxicity as an endpoint? Obviously, the endpoints in a trial depend on the trial itself—phase I studies are geared towards toxicity evaluations, phase II towards effectiveness, and phase III towards survival.

It occurred to him—maybe the endpoints should not be determined solely by a group of investigators in a conference hall in San Diego. Maybe it's time to have this dialogue with the public—beyond the National Cancer Institute, beyond the American Cancer Society, beyond ASCO. Maybe they should engage the public into a discussion of what endpoint is "meaningful."

I liked the reply to his commentary by Dr. Rick Boulay. Adopting the same endpoints in molecular medicine as they have in the era of cytotoxic chemotherapy, may lead to erroneous conclusions. Reports in the New England Journal of Medicine from ICON 7 and GOG on the use of Avastin (bevacizumab) in frontline ovarian carcinoma have shown modest improvement in disease free interval while overall survival may not be increased.

Further teasing out the data shows that shortly after Avastin is discontinued, patients tend to recur. Are they using this drug wrong? Are the expectations of improvement in efficacy despite discontinuation of a biological agent too high?


Biologic therapy is on the ascendancy. The possibility of eradicating cancer by selective destruction of tumor blood vessels may represent an attractive therapeutic avenue. It's going to take "combination" antivascular therapy though, to make a difference. The AngioRx Assay is the most promising technology on the therapeutic horizon. And on top of that, putting a vascular disrupting agent in combination with an anti-angiogenesis agent.


Posts: 1223
Joined: Jun 2008

thanks for posting this. I have a question. What would be considered to be a "disrupting agent" to combine with anti-angiogenesis agent? I assume taxol or taxotere would NOT be that option (they are in themselves angiogenesis agents,right? I assume the answer would be some kind of a platinum or something other than a taxan?

This is a crucial bit of information, as I am starting avastin very soon, and my doc in Tokyo recommended adding only taxotere. My gut tells me it needs to be a platinum if it is a single agent added to the avastn.

Any thoughts?


gdpawel's picture
Posts: 538
Joined: May 2001


The only way for taxol or taxotere (or most any other cytotoxic agent like a platinum) to act like an anti-agiogenesis agent, it must be given as a "low-dose" chemotherapy protocol. When administering these drugs, the endothelial cells (involved in angiogenesis) are the first in the tumor to undergo cell death (apoptosis). The anti-angiogenic effect does not translate into a significant therapeutic benefit because the damage to the vasculature of the tumor can be largely repaired during the long rest and recovery periods between successive cycles of therapy.

The more frequent, lower-dose therapy can have an impressive anti-angiogenic and anti-tumor effects. Blood vessel cells are less likely than tumor cells to become resistant to chemotherapy, so if cancer cells become drug resistant, these medicines should still be able to shrink tumors by destroying their blood supply.


The paper I did (Who Needs Taxol?) in regards to dissemination after taxane-based chemotherapy, had to do with "dose-intense" therapy, not "low-dose" therapy. Remember, "dose-dense" is not the same as "low-dose."


A vascular disrupting agent goes after the abnormal (sensitive) part of an established tumor vasculature (abnormal vs healthy). It already is highly fractured and architecturally disorder with fragile, immature blood vessels which lack external smoother muscle and pericyte support. The result is block blood flow to the tumor. With Avastin, tumor starves from the outside. With Zybrestat, the tumor starves from within.

Angiogenesis inhibitors target the formation of new blood vessels. Vascular disrupting agents target endothelial cells and pericytes (connective tissue) of the already established tumor vasculature, resulting in tumor ischemia (inadequate blood supply) and necrosis (death). So the thought here is to go after the formation of new blood vessels (Avastin) and go after the already established tumor vasculature (Zybrestat).

Pericytes are in capillaries which regulate the blood-brain barrier.

When tissues (normal and neoplastic) increase in size, they require the formation of microcapillaries (angiogenesis) to provide nourishment to sustain their growth. Constituents of these microcapillaries include most prominently endothelial cells, but also associated mesenchymal cells, fibroblasts, smooth muscle cells, and pericytes. Angiogenesis is important in normal processes, such as wound healing, but also in diseases such as cancer, psoriasis, diabetes, rheumatoid arthritis, and age-related macular degeneration.

Mature vessels are coated by pericytes, mesenchymal-like cells, which support these vessels and enable improved blood flow. The immature vessels are only tubes of endothelial cells, which are more “leaky” than the mature ones, and conduct oxygen and nutrients less effectively to the tumor.

In order to use Zybrestat (in the US) for "off-label" use with Avastin, it would have to obtain US FDA approval in at least one successful clinical trial or for you to be in one of those clinical trials. The only clinical trial that I know of with Zybrestat is stage II in lung cancer.


gdpawel's picture
Posts: 538
Joined: May 2001

Dr. Robert Nagourney, of Rational Therapeutics, posted on his blog, his take on the July 2011 meeting of the Ovarian Cancer National Alliance (OCNA), which included a lecture by John Hays, MD, from the National Cancer Institute (NCI), entitled “Decision time: what is the right choice of chemotherapeutic agent(s)?

Dr. Hays, part of the molecular signaling section at the NCI, reviewed literature on the topic. He described the need for prospective clinical trials to validate retrospective and in vitro results.He then examined data from three technologies, the Oncotech extreme drug resistance test, Precision Therapeutics ChemoFX test and the ATP-based chemosensitivity test.

Dr. Nagourney found it odd that Dr. Hays spent time examining the EDR (extreme drug resistance) technology of Oncotech in as much as it is no longer offered and reflects proliferation-based studies, which have since largely been discredited.


The ATP assay was reviewed using the results of a study published by Dr. Ian Cree in which 180 patients received either assay-directed (ATP) or physician's choice. This study actually provided an improvement for patients who received the ATP-based treatment but failed to achieve significance. It failed largely because it was underpowered.

But this reflected a more concerning aspect of the study. It seems that the “physician's choice” arm increasingly applied the best drug regimens developed in Dr. Cree’s own laboratory. As the trial continued to accrue, an increasing proportion of patients received Gemzar (gemcitabine)-based doublets (which were very new at the time) based upon Dr. Cree’s observation of activity for these novel combinations.

This had the uncomfortable effect of forcing Dr. Cree to compete with himself. Had Dr. Hays been truly interested in examining this study as Dr. Nagourney has, he might have noticed the good control group response rate partly reflected the application of Dr. Cree’s’ own observations.

Indeed, when during Dr. Nagourney's many attempts to conduct a prospective study with the GOG, he was confronted with a study design similar to Dr. Cree’s, (e.g. they could incorporate any treatment they chose, including those that he developed), his statistician demanded that he forego the pleasure, as he could see only too well that the trial had become impossible to power. There was no true control arm for statistical comparison.


The final portion of Dr. Hays’ presentation was the ChemoFX assay. This technology propagates tissue biopsies to confluence and then conducts measurement of drug-induced cell death.

With substantial funding largely provided by venture capital, Precision Therapeutics has leapt into the GOG with a series of trials. Should this hybrid technology fail to provide prospective results that meet significance, it will be a damaging blow to this unfairly maligned area of investigation.

While Dr. Nagourney wishes the ChemoFX investigators luck, a failure on their part could be harmful to the field. Their reliance on propagated, sub-cultured tissues grown to monoculture has been a concern to Dr. Nagourney since they first arose in the last few years as participants in the field.


According to Dr. Nagourney, what is interesting in Dr. Hays’ review is not so much what he said, but what he didn’t say.

First, he did not mention the seminal work of Dr. Larry Weisenthal, of Weisenthal Cancer Group, a pioneer in the field.

Second, he did not describe the nearly 2,000 retrospective, yet statistically significant correlations in the literature in a wide variety of diseases. He neglected to mention that one of the most widely used regimens for breast and ovarian cancer was developed using the same human tumor culture analyses that he decries. If he actually treats patients, he no doubt uses the cisplatin gemcitabine doublets developed using one of these platforms.

Finally, Dr. Hays has failed evidence-based medicine 101. He has forgotten that in life-threatening illnesses where prospective clinical trial data is not available, in accordance with the dictates of evidence-based medicine, one should use the best available data to guide treatments.

There is a wealth of data supporting laboratory based drug selection. Presentations like that described do not add to the discourse.


gdpawel's picture
Posts: 538
Joined: May 2001

It was hoped that something like the Arienti, et al study would be proposed at the semi-annual GOG meeting. Perhaps a good three-armed clinical trial: physicians' choice vs molecular profiling vs functional profiling?

It looks like Dr. Larry Weisenthal of the Weisenthal Cancer Group has given some information in relation to this.


gdpawel's picture
Posts: 538
Joined: May 2001

Susan Jenks

Imagine taking a snapshot of a tumor in action as it overwhelms healthy cells to survive—and then using that information in the clinic to monitor which drugs best kill cancer cells. In vitro chemosensitivity assays promise to do that, using proprietary screening tests for clues about each individual’s cancer.

But more than two decades after these assays debuted, oncologists remain divided: those who support such laboratory analyses and those who maintain that they unduly raise patients’ hopes, often at considerable expense, while not improving survival. “Predicting what doesn’t work [as opposed to what will] is not seen by most as an advance,” said Jeffrey Abrams, M.D., associate director of the National Cancer Institute’s Cancer Therapy and Evaluation Program.

Last year, the American Society of Clinical Oncology (ASCO) performed a literature review of data published between December 1, 2003, and May 31, 2010, on chemosensitivity testing. ASCO concluded, as it had in 2004, that this “in vitro analytic strategy has potential importance” but should be confined to patients participating in clinical trials. The National Cancer Comprehensive Network holds a similar view.

“In vitro testing is an old oncologist’s dream,” said Alain Hendlisz, M.D., chief of gastroenterology in the medical oncology clinic at the Jules Bordet Institute in Brussels, where researchers are using metabolic imaging to evaluate colorectal cancer treatments after an initial round of therapy.

“In an ideal world, we should know what will be the effect of a treatment before its very beginning. That’s the purpose of in vitro testing, and that’s the philosophical reason why people are attracted by tests with a high positive predictive value,” Hendlisz wrote in an e-mail. “However, that’s a philosophical point of view, not a scientific one.”

Still, in certain circumstances, where no standard of care exists, chemosensitivity testing may be useful to determine how best to treat some patients, according to Robert Holloway, M.D., co–medical director of Florida Hospital’s gynecologic oncology program in Orlando. Holloway routinely recommends such tests for his ovarian cancer patients, at a cost of $3,500–$4,000.

Although insurance coverage tends to be spotty, Medicare does cover the cost for ovarian cancers, because these tumors account for most of chemosensitivity testing’s use today, Holloway said.

His decision to use the assays is based on “strong retrospective data” in at least two recent studies in ovarian cancers, including one by Julian Schink, M.D., and Larry J. Copeland, M.D., of the Comprehensive Cancer Center of Northwestern University, showing that chemosensitivity assay testing closely predicted patients’ responses and "accurately predicted progression-free and overall survival.” The study was published in the September 2011 Journal of the National Comprehensive Cancer Network.

Also, the development in recent years of several drugs for ovarian cancer, which has a high recurrence rate, makes selecting the right drug problematic for clinicians and the field ripe for this type of testing, Holloway said.

“Through chemosensitivity testing, we try to narrow down the selection,” he said. “Our feeling is it can’t be any worse than random selection, and there is logic behind it. I would caution, however, it’s not something that makes an absolute decision for you. You still weigh toxicities and other risks, such as how readily accessible the tissue is, but it provides another piece of information.”

Laura Shawver, Ph.D., an ovarian cancer survivor and founder of a nonprofit for ovarian cancer patients called the Clearity Foundation in San Diego, takes a nuanced view of chemosensitivity’s clinical value for ovarian and other cancers. Her main concern: the lack of standardization in these tests.

“Unfortunately, with chemosensitivity testing, it’s not done in a standardized format,” she said. “Some labs use apoptosis, some use cell growth, and still others use 3-D assays where they grow cells in clusters so the assay more closely resembles a tumor.” Even how a tumor is removed from the body can adversely affect the reliability of an assay’s results, she added.

Shawver favors genomic analysis, and her nonprofit helps women diagnosed with ovarian cancer gain access to molecular profiling to find the best treatment options at recurrence. She views genomics as potentially “a way to supplement or even replace chemotherapy” someday and a technique far better poised to capture tumors’ heterogeneity than chemosensitivity testing.

“People need to realize tumors change over time; they are highly mutable," she said. "There's heterogeneity at diagnosis, so chemosensitivity testing will test a portion of those tumor cells only, and not wipe them all out."

But whenever chemosensitivity assays are available (Shawver does not recommend them), she does compare their results with molecular profiling. “We should think of these assays as a way of prioritizing drugs, not as a yes or no,” she said.

David Alberts, M.D., an Arizona Board of Regents professor and director of the University of Arizona Cancer Center, takes an equally pragmatic view, suggesting a limited role for chemosensitivity testing, at least in patients with uncommon tumors, where no "gold standard" of treatment exists. In such cases, he said, "I want as much information as I can get."

And though he concedes a lack of clinical data, Alberts also sees a similar lack of “adequate clinical testing” of genomic assays. “These tests don’t tell you what proteins are expressed . . . or about drug interactions during treatment.”

In several cases where patients received dire diagnoses, Alberts sent their tumor samples for functional/chemosensitivity testing and to an international genomics center in Tucson for expression profiling. The two assays yielded similar results and helped guide the patients' successful treatments. "There's no question from my point of view that there's a national need for randomized well-designed clinical studies, comparing all these assays," Alberts said.

Jonathan Lancaster, M.D., Ph.D., chairman of the department of women’s oncology at Moffitt Cancer Center in Tampa, Fla., agreed. But with the expense and the difficult risk–reward balance of proving that an assay works, he said, such studies will probably not be conducted anytime soon. At Moffitt, he said, oncologists rarely use chemosensitivity testing, except when patients’ cancers prove so intractable that, like Alberts, “we feel as if we’re stumbling in the dark.”

Over the years, chemosensitivity tests have become more refined, evolving from detecting only drug resistance to more sophisticated biomarkers that yield better information about a tumor’s biology, Lancaster said. But in his view, two fundamental questions remain: (1) the science of prediction, or whether a bio- marker predicts what it’s supposed to predict, and (2) what happens to a patient when you use the information gleaned from these tests.

“[The latter question] has been the real stumbling block,” Lancaster said. “Even if a test predicts a response, you can’t cure that cancer.”

No studies have shown that chemosensitivity testing improves overall survival. The 2011 literature review by ASCO’s working group found that of 11,313 new articles on chemosensitivity testing within the 7-year window assessed, only 21 met predefined inclusion criteria. Of these, five were randomized clinical trials, none of which supported chemosensitivity testing in oncology practice, given the few patients involved.

Such studies are incredibly difficult to do, Lancaster said. Moreover, many tests in medicine today don’t categorically improve survival, he said, such as weekly blood count monitoring after chemotherapy or even some widely used cancer drugs. “We can delay growth of a tumor or increase the time of stable disease, but does it improve survival? Are we holding these assays to a higher standard than we do other tests?”

One practitioner, Robert Nagourney, M.D., Ph.D., medical and laboratory director of a company in California, who has been doing chemosensitivity testing for more than two decades, thinks so. From his perspective, the scientific elegance of genomics testing has obscured research interest in chemosensitivity assays, which were "not well done for a long time, so they were dismissed."

But new tests can better capture the microenvironment of tumor cells inside the body, he contends, enabling technicians to isolate cancer cells into clusters that include even some of their vasculature.

Still, doctors generally prefer to follow treatment results from clinical trials rather than from the laboratory. According to Richard Levine, M.D., president of Space Coast Cancer Center in Titusville, Fla., and a member of ASCO’s National Quality Care Committee, “If you have a standard regimen for a particular cancer, but chemosensitivity testing says don’t give a certain drug, there’s no article that shows a patient will do better. We have to go with the known data."

Although genomic testing already carries storied data (including advances with Her2/neu, ALK inhibitors, and BCR– ABL gene mutations in chronic myeloid leukemia) that have influenced patient care, chemosensitivity assays have not yet attained the same level of success, Levine said.

But a recent ASCO initiative called CancerLinQ might change that. The mul- timillion-dollar massive electronic database will record, track, and collate chemosensitivity and molecular profiling testing results for clinicians to access.

“This system should help guide treatment and help us be current,” Levine said. “It will have all the toxicities, response rates, and survival, so doctors can mine the data for the best care for their patients.”

Oxford University Press 2012. DOI:10.1093/jnci/djs499

Randomized Clinical Trial Paradigm


Subscribe to Comments for "GOG Meeting"