Avastin maker Genentech argues that PFS should be the primary endpoint, while FDA believes that the drug’s failure to significantly improve OS is the key issue. [James Steidl - Fotolia.com]
FDA’s success in removing Avastin’s metastatic breast cancer (MBC) indication has sparked a longstanding debate in drug development circles: What is the best proof that a drug works and should be approved? A panel discussion held November 29 in Cambridge, MA, debated whether the primary endpoint for regulatory sanction should remain overall survival (OS) or if it should change to progression-free survival (PFS)?
In Avastin’s case, a divided FDA initially gave the drug accelerated approval, in combination with paclitaxel, for MBC in 2008. The decision was based on a clinical study showing a 5.5-month increase in PFS to 11.3 months. Starting last year, though, FDA’s Center for Drug Evaluation and Research (CDER) began recommending that the MBC indication for Avastin be withdrawn.
CDER pointed to new studies, one of which showed a difference in overall survival for patients with MBC of less than two months—from 24.8 months to 26.5 months. “The addition of Avastin to paclitaxel resulted in an improvement in PFS with no significant improvement in OS,” CDER declared in an April 7 filing.
Avastin maker Genentech and parent Roche argued that PFS should be the primary endpoint, because it could be objectively measured and because the EU recognized it as an endpoint in oncology clinical studies. FDA countered that the primary endpoint should instead be overall survival, while PFS should continue as a surrogate endpoint of clinical benefit. Additionally, FDA said Avastin’s PFS improvement was not worth the risk of associated side effects like severe high blood pressure, hemorrhaging, heart attack or failure, and the development of perforations in the nose, stomach, and intestines.
Roche still hopes to reinstate the indication for previously untreated MBC through a Phase III study. The study will in part examine a potential biomarker that the company says may help identify patients likely to derive substantial benefit from the drug.
Raising the Bar to OS
During the discussion held on November 29, called “Surrogates or Survival: A Discussion on Endpoints in Multiple Myeloma Clinical Trials,” moderator Anne Quinn Young recalled how cancer treatment has changed over the past 10 years. A decade ago, she said, FDA was willing to accept PFS as an endpoint because so many cancers had a poor prognosis that to expect OS improvement wasn’t relevant.
That has changed as new cancer treatments have led to patients surviving longer and in better shape, continued Quinn Young, vp of strategic alliances for the Multiple Myeloma Research Foundation (MMRF). “Patients and clinicians have raised the bar,” she said. “They’re not putting up with high toxicities any more.
“There’s kind of a shift toward the view that an advantage in progression-free survival is one thing, but now we have agents that are demonstrating an advantage in overall survival. So perhaps we should start looking at the way patients are treated in a little bit different way.”
Establishing Treatment-Specific Endpoints
Both PFS and OS are appropriate endpoints to define clinical benefits for patients with metastatic cancer, Axel Hoos, M.D., Ph.D., co-chairman of the Cancer Research Institute Cancer Immunotherapy Consortium (CIC), told GEN: “Ultimately, to define patient benefit, the key question is how great is the benefit of either of the two endpoints to say if it’s clinically meaningful or not.”
Additionally, Dr. Hoos noted, “it depends on some of the agents that are being investigated, so the type of therapy and then what endpoint is more useful.” For example, “if immunotherapies are less prone to deliver conventional responses, and this is an endpoint, then that is only incompletely reflecting the clinical benefit that an immunotherapy can provide,” Dr. Hoos added. “It doesn’t mean that PFS is not a useful endpoint. It still is. But it doesn’t tell the whole story.”
Dr. Hoos also serves as medical lead in immunology/oncology at Bristol-Myers Squibb (BMS). “For the immunotherapies that we’ve investigated so far,” he commented, “we believe survival is the better endpoint because it encompasses the full spectrum of clinical benefit.”
Dr. Hoos noted that two cancer therapies approved in recent years have shown clear survival benefits—Dendreon’s Provenge vaccine and BMS’ Yervoy (ipilimumab) antibody for melanoma: “For Provenge, you have no PFS improvement at all. For ipilimumab, there is a PFS improvement, but it is smaller than the survival improvement. So in both cases, I would say, it’s very clear that survival is the right endpoint.”
Translating Clinical Endpoints to Treatment Benefit
The fact that both endpoints have their merits under varied conditions suggests that their value to patients may indeed, as the panelists concurred, depend on the individual. Noopur Raje, M.D., director of the multiple myeloma program at Massachusetts General Hospital and one of the panelists, said decisions on treatments and their endpoints need to be tailored to individual patients based on factors that include age, mobility, and income.
“Ultimately, of course, the absolute endpoint should be: Our patients should live longer, but not just longer; they should live better with myeloma.” As Dr. Raje correctly noted in relation to multiple myeloma but which applies to other forms of cancer, there’s a lot of heterogeneity within diseases—heterogeneity that researchers are only now beginning to take a serious look at.
“The way we treat myeloma today is one size fits all. And that’s not true at all,” Dr. Raje said. “Absolutely, I think the next step forward is to better re-stratify, better understand the genetics of patients’ myeloma, and to do targeted treatment. And that’s the level of understanding we need to get to.”