New therapies improving survival in patients with glioblastoma

Experts agreed that this is a groundbreaking era for the treatment of gliomas.

Although gliomas and glioblastomas are not the most commonly occurring cancers — there are about 10,000 new cases in the United States each year — they are among the most deadly. The median survival of people diagnosed with these cancers is between 12 and 15 months.

However, a recent article in The New York Times estimated that the percentage of patients who survive two years from diagnosis of glioblastoma has more than tripled in the last five years, from 8% to 25%, largely because of the use of temozolomide (Temodar, Schering Plough) plus radiation as part of treatment regimens. That, in addition to progress made with bevacizumab (Avastin, Genentech), an antiangiogenic VEGF inhibitor, means that research and treatment options from gliomas are more exciting now than they have been for the last 20 years, according to Howard Fine, MD.

“I’ve been treating glioblastoma for about 22 years. I’ve taken care of more than 20,000 patients,” said Fine, chief of the neuro-oncology branch at the National Cancer Institute’s Center for Cancer Research and of the National Institute of Neurologic Disorders and Stroke. “The kinds of things we’ve seen in the clinic in the last four years blows away anything I saw in the previous 18 years of my career.”

Bevacizumab was granted accelerated approval for the treatment of patients with relapsed glioblastoma in May. In addition to gains in two-year survival seen with the use of temozolomide, results from several phase-2 studies of bevacizumab have shown improvements in PFS and objective response rates.

Howard Fine, MD, chief of the neuro-oncology branch at the NCI’s Center for Cancer Research, has treated glioblastoma for 22 years. Photo by Dupont Photographers

Traditionally, six-month PFS survival in relapsed or progressive glioblastoma is about 9% to 21%, and objective response is less than 10%. According to results published in the Journal of Clinical Oncology in October, patients with recurrent glioblastoma had an estimated six-month PFS of 42.6% while assigned to bevacizumab alone and 50.3% while assigned to bevacizumab plus irinotecan.

Henry S. Friedman, MD, and colleagues who conducted the study said the PFS exceeded the expected 15% rate assumed for salvage chemotherapy and irinotecan alone. Furthermore, the objective response rate was 28.2% in the bevacizumab group and 37.8% in the bevacizumab plus irinotecan group.

Despite these promising results, much remains to be seen regarding treatment with bevacizumab, according to Patrick Y. Wen, MD, associate professor of neurology at Harvard Medical School. Wen published an editorial in Expert Review of Anticancer Therapy in September. In it, he wrote that bevacizumab’s cost is an economic issue that needs to be resolved, that although bevecizumab clearly has benefits, they may be transient, and that the improved PFS associated with the drug may not always translate into improved OS.

Bevacizumab and similar drugs clearly shrink tumors, Wen told HemOnc Today, but they have not been shown to destroy tumors. “Trying to understand the mechanisms of resistance to improve on these therapies has been a real challenge,” he said. “That’s probably the hottest area in neuro-oncology right now.”

Immunology

In addition to the more standard temozolomide and the newly approved bevacizumab, there are several other therapies, such as vaccines and targeted molecular therapies, which are giving hope to the experts who spoke to HemOnc Today. Friedman, the James B. Powell Jr. Professor of Neuro-Oncology at Duke University, said medical science could be on the cusp of a cure.

“I believe that temozolomide made a small step forward. I believe bevacizumab will make an even bigger step forward, and I believe vaccines will make another step forward,” Friedman said. “There is an ever increasing minority of patients who appear to get cured of this disease. Within the next five years, I hope to see a major increase in survival that is bevacizumab-mediated, and maybe vaccine-mediated as well.”

In 2008, Wheeler and colleagues published results in Cancer Research from a phase-2 trial of a glioblastoma multiforme vaccine. Thirty-four patients with glioblastoma and 10 healthy participants were assigned to vaccination with autologous tumor lysate-pulsed dendritic cells at Cedars-Sinai Medical Center. Researchers said they hoped that the vaccine would stimulate the host immune system to mount a specific cytotoxic T-lymphocyte response against mesothelioma tumor cells, resulting in tumor cell lysis. Primary endpoints were time to progression and time to survival.

Time to survival in patients who responded to the vaccine was 642 days compared with 430 days in nonresponders. Time to progression was 308 days in vaccine responders compared with 167 days in nonresponders. The researchers found that time to survival and time to progression in all vaccinated patients with glioblastoma compared favorably to patients treated with nonvaccine therapies at Cedars-Sinai during the course of the trial.

As in earlier studies, the vaccine was found to be safe and well-tolerated; there were no grade-3 or grade-4 adverse reactions reported.

Henry S. Friedman

“Our findings validate post-vaccine chemosensitization of glioblastoma multiforme by vaccine-induced T-cell responses,” Wheeler and colleages wrote. “It should be noted that these findings, although promising, require verification in larger groups of patients treated in randomized trials. Nevertheless, based on this promise, it is anticipated that screening vaccine candidates for markers predicting vaccine responsiveness, combining vaccination with recently improved chemotherapeutic regimens, and rendering vaccines accessible to more patients may help maximize vaccine-mediated clinical benefits for glioblastoma multiforme.”

Among the current studies into immunotherapies for glioblastoma, researchers at The University of Texas M.D. Anderson Cancer Center are currently conducting a trial looking at WP1066, an orally delivered STAT3 inhibitor. The ultimate goal is to determine if inhibition of STAT3 might directly kill glioblastoma cells and stimulate the immune system to attack those tumor cells.

Friedman said there are several vaccines undergoing clinical trial. Pfizer’s CDX-110, a glioma-associated antigen peptide-pulsed autologous dendritic cell vaccine, is being researched by Linda M. Liau, MD, PhD, at Jonsson Comprehensive Cancer Center at UCLA. Another is the attenuated cytomegalovirus vaccine being developed at Duke. Friedman said he hopes that, because vaccines represent a different strategy for attacking glioblastoma, at least one vaccine will eventually enter clinical practice.

Wai-Kwan Alfred Yung, MD, chair of neuro-oncology at M.D. Anderson Cancer Center, said he is sure vaccines will prove a viable strategy in the near future.

“Immunotherapy should be on our radar screen. We should probably put more effort and more science into exploring vaccines,” he said. “There are many ongoing trials looking into vaccines, but the early results don’t show any one vaccine standing out.”

Molecularly targeted therapies

Fine said that the most hope for a cure may lie in molecularly targeted agents. Molecularly targeted therapies reflect the reality that cancer cells are different than normal cells due to mutation, so this class of drugs is selective against the cancer cell and not the normal cell.

“That theory has been around for a long time, but for the first time we’re actually seeing evidence, not just in our mice and our rat studies but in our patient studies, that these kinds of agents can actually change the course of this disease and cause tumors to regress in a way I’ve never seen before,” Fine said. “They represent a more rational way of treating cancer as opposed to in the past and still today, where standard treatment for cancer is made up of poisons — be it standard cytotoxic agents, chemotherapy agents or radiation.”

The VEGF inhibitor bevacizumab seems to be the best-known and most-tested treatment of this group. Phase-2 studies have shown that alone and in combination with irinotecan, bevacizumab was effective and well-tolerated in patients with glioblastomas.

The epidermal growth factor receptor is amplified in roughly 50% of glioblastomas and over-expressed in many malignant gliomas regardless of amplification status, and results have shown that the presence of EGFRvIII is an independent prognostic factor for poor survival. This would seem to suggest a role for EGFR in glioma pathogenesis and offer an obvious rationale for developing therapies that target EGFR.

However, Sathornsumetee and colleagues concluded after studying the existing literature that the EGFR kinase inhibitor gefitinib (Iressa, AstraZeneca) had minimal impact on radiographic response. Results of studies evaluating another EGFR kinase inhibitor, erlotinib (Tarceva, OSI), showed response rates from 6% to 25%, but neither drug had a clear impact on survival.

“We’ve done trials for erlotinib in glioma and there’s about a 10% response rate. But just having the presence of the EGFR or having an active EGFR on the tumor cells is not enough to predict who those 10% of responders are. It’s much, much more complex than that,” Fine said. “We’re not yet at the point where we understand that complexity. It’s more than just finding the right biomarker to select the right patients for the right drug. Finding that biomarker that identifies what will predict for response to these agents is still a tremendous work in progress.”

Molecular pathways

The other target for molecular therapies is finding treatments that attack the molecular pathways that sustain tumors. Wen said that because tumor stem cells are of particular importance in glioblastomas, understanding the pathways that drive those stem cells is crucial to treating this disease.

Currently, trials are exploring inhibitors of the Notch pathway and the hedgehog pathway, Wen said. “There’s a transcription factor called OLIG2 that’s thought to be very important in glioma stem cells. Finding ways to knock down OLIG2 is an important therapeutic strategy, but that’s not clinically available right now.”

According to Fine, targeting pathways represents a new way of treating cancer.

“There’s a whole set of developmental pathways that our colleagues in developmental biology have known about for years but that were never in the repertoire of cancer biologists and so were never in the repertoire of pharmaceutical companies that think about targeting cancer pathways,” he said. “All of a sudden, there’s this great convergence of developmental and stem cell biology pathways with cancer pathways. It opens up a whole new series of molecular targets such as Wnt, sonic hedgehog and Notch.”

Perifosine (Keryx Biopharmaceuticals), an oral ATK inhibitor, is currently under evaluation for use in patients with malignant gliomas. Sirolimus (Rapamycin, Wyeth) and its synthesized analogs, temsirolimus (Torisel, Wyeth), everolimus (Afinitor, Novartis), and AP23573 (Ariad Pharmaceuticals) have been evaluated in clinical trials of malignant gliomas as inhibitors for mTOR, a serine/threonine kinase downstream from AKT.

“Preclinical studies demonstrated that inhibition of mTOR can stimulate the kinase activity of its immediate upstream effector, AKT, which may decrease the antitumor efficacy,” Sathornsumetee and colleagues wrote in Cancer in 2007. “PI-103, a novel inhibitor of both PI3K and mTOR, has shown promising activity in both in vitro and in vivo models of malignant gliomas, partly because of blocking activated PI3K/AKT induced by mTOR inhibition.”

Sathornsumetee and colleagues wrote that studies conducted by the North American Brain Tumor Consortium and the North Central Cancer Treatment Group have shown evidence that temsirolimus is associated with radiographic improvement, but that has not translated into improved survival.

“Targeting molecular pathways is very promising because we now have a lot more information on each tumor that might be utilizing the growth signal,” Yung said. “We will increase our efficacy when we learn how to classify them better so we can cater a drug to a specific group of tumors.”

Designing effective trials

The experts who spoke to HemOnc Today all cited the difficulty of designing randomized, controlled trials to investigate glioma treatments. Because the disease is highly heterogeneous, each patient in many ways represents a separate trial. Fine noted that it is becoming a priority to consider new paradigms of clinical trial design to address the tremendous heterogeneous nature of this disease so that each clinical trial for a glioblastoma therapy does not take decades or millions of dollars to conduct.

“If it was simple, all you’d do is screen patients for, say mutation B, and design a trial of patients who just have mutation B,” he said. “The problem is it’s not all that simple. These molecularly targeted agents don’t always just identify mutation B. They may be effective against mutation B, but only when you have mutations A and G, but normal gene X. It’s a highly complex system and one we don’t yet understand.

“Unfortunately, it may turn out that we may have to do trials the old fashioned way and take all comers, then retrospectively look at that 10% of patients who respond and find out what was unique about them and their tumors before you can identify the probable predictors of responsiveness, and then design a prospective trial with people who have just those markers in order to prospectively confirm the observation,” Fine said.

Eventually, study populations will be tightly screened, Friedman said. Physicians will identify certain key pathways and only enroll patients who have abnormalities.

“Study populations will be much more homogeneous,” he said. “That’s the hope of the future. That’s not the future right now.”

Yung said that all solid tumors are heterogeneous, yet researchers have figured out how to design effective trials to evaluate treatments for lung cancer and breast cancer.

“No two glioblastomas are the same, but give me 100 glioblastomas and there may be three or four different groups and each group may respond to a different targeting agent,” he said. “We have to learn how to match the signal from the tumor to the signaling agent we need to use.”

Ultimately, the experts agreed that, for the first time in a long time, the trends in glioblastoma treatment are moving in the right direction.

“We have made some small steps. I am very happy to see that we’re having more long-term survivors. We’re bending the tail end of the curve more and more,” Yung said. “We’re close to moving the median survival in a substantial way. I don’t know whether we can use the word breakthrough, but I think we’re on the verge of making some major improvements.” – by Jason Harris

Is there an advantage to administering bevacizumab directly to brain tumors using microcatheters?

For more information:

• Amstrong TS. Oncology Nursing Forum. 2009;doi:10.1188/09.ONF.E232-E240.
• Bredel M. Lancet Oncol. 2009;doi:10.1016/S1470-2045(09)70192-5.
• Fine HA. J Clin Oncol. 2004;doi:10.1200/JCO.2004.06.927.
• Friedman HS. J Clin Oncol. 2009;doi:10.1200/JCO.2008.19.8721.
• Grady D. Breaching a Barrier to Fight Brain Cancer. New York Times Online. Available at: http://www.nytimes.com/2009/11/17/health/17tumor.html?_r=1&em. Accessed Nov. 17, 2009.
• Sathornsumetee S. Cancer. 2007;doi:10.1002/cncr.22741.
• Steiner HH. J Clin Oncol. 2004:doi:10.1200/JCO.2004.09.038.
• Stupp R. Lancet Oncol. 2009;doi:10.1016/S1470-2045(09)70025-7.
• Wen PY. N Engl J Med. 2008;359:492-507.
• Wheeler CJ. Cancer Res. 2008;doi:10.1158/008-5472.CAN-07-5973.