One-quarter of patients with metastatic pancreatic cancer who received the Hedgehog inhibitor vismodegib (formerly called GDC-0449) and gemcitabine together had a confirmed partial response in a pilot trial, researchers reported at the American Association for Cancer Research's Pancreatic Cancer: Progress and Challenges conference. Additionally, the investigators, led by Edward J. Kim, MD, PhD, a medical oncologist and Clinical Lecturer at the University of Michigan Comprehensive Cancer Center, say they think they have identified a biomarker that predicts response to the combination.
"The response rate was very high," said Daniel Von Hoff, MD, Physician-In-Chief at the Translational Genomic Research Institute (TGen) and Co-Chair of the conference. The typical response rate for gemcitabine monotherapy is about five percent. "Here it was 25 percent, and that was beyond what normally one would expect, so I am very interested to see the final results of this," continued Von Hoff, who was not involved in the trial.
'Eyes Wide Open'
Cancer biologists have shown that Hedgehog signaling plays a critical role in pancreatic cancer. For example, David A. Tuveson, MD, PhD, Senior Group Leader at the Cambridge Research Institute in the UK, and colleagues showed that Hedgehog signaling contributes to the dense stroma characteristic of pancreatic cancer. Moreover, when they inhibited the pathway in a genetic mouse model of the disease, they saw a reduction in intratumoral pressure and better penetration of anti-tumor drugs (OT, 7/10/09).
More recently, Kim and colleagues reported that Hedgehog signaling, which is absent in healthy adults, is turned on in pancreatic cancer stem cells.
Despite this strong biological rationale for Hedgehog inhibition in pancreatic cancer, an early randomized trial with the Hedgehog inhibitor saridegib (formerly called IPI-926) plus gemcitabine showed no benefit for the combination compared with placebo plus gemcitabine and was stopped early due to futility.
Given those conflicting data, Kim's team designed the current trial in such a way that they could assess vismodegib's impact on the tumor biology as well as on the patients. Patients in the trial received vismodegib monotherapy for four weeks and then started on vismodegib plus gemcitabine. The investigators biopsied liver metastases from each patient at baseline and after three weeks of vismodegib monotherapy.
"A key to the success of any new targeted drug or treatment is to identify predictors of who will benefit," Kim said during an AACR news conference in which he presented the preliminary trial results. "Our sequential treatment design allowed us to study tumor tissue before and after GDC-0449 therapy. We could see the impact of GDC-0449 alone on Hedgehog signaling, on tumor stroma, and on CSCs [cancer stem cells]."
The team is still analyzing the paired biopsies in terms of the drug's effect on tumor stroma and cancer stem cells, but they can already see that shutting off signaling is an important factor in response.
"We determined that pretreatment Sonic Hedgehog expression level is the best predictor of response," Kim said, noting that in addition to the five patients with partial responses, another five individuals had stable disease on therapy at three months.
Tuveson, also a co-chair of the meeting, pointed out during the news conference that this is the first trial in pancreatic cancer to use paired biopsies, and that this approach was particularly important after the previous trial's failure.
"We are going into this trial with our eyes wide open," he said. "Potentially one of the most important things we are going to learn from this trial is that there is a subset of patients that potentially benefit. We have been taught that over and over about targeted therapies in other cancer types. [horizontal ellipsis]When we get to the randomized trial stage, it may be in a setting where we can select for a population that have a chance of benefit; that may be the first time in pancreatic cancer in which we've done such a trial."
When asked to compare the two Hedgehog inhibitors, vismodegib and saridegib, Von Hoff declined to give many details because he has tested both agents in clinical trials-and signed confidentiality agreements. But he did say that these two agents, as well as other Hedgehog inhibitors, are not all the same. Some of the inhibitors are chemically similar to one another, while others fall into totally different classes. And with regard to these two specifically, it is known that they differ in several ways, including the blood levels that can be achieved. Therefore, he is not surprised, he said, that one might work in one disease setting while the other one would not.
He cautions, though, that before anyone makes too much of the current results, promising as they are, the combination will need to be tested in a randomized trial. The good news, though, he said, is that these Hedgehog inhibitors do not add any kind of substantial toxicity whatsoever."