By Turna Ray
Following the accelerated approval of Vectibix in the US in 2006, follow-on studies have failed to show that metastatic colorectal cancer patients on the drug live longer than those receiving standard chemotherapies. However, Amgen this week said it is in the process of conducting a study to try to prove that Vectibix-treated chemorefractory patients whose tumors express the wild-type KRAS gene do survive longer than patients receiving just chemotherapy.
The protocol for this trial is currently being reviewed by the US Food and Drug Administration. According to Amgen, this study is necessary to confirm Vectibix's clinical benefit in metastatic colorectal cancer patients and to translate Vetibix's accelerated approval status to full regulatory approval with the FDA.
When Amgen presented its plan to conduct this study before FDA's Oncology Drugs Advisory Committee this week, panelists were doubtful that the company will be able to demonstrate overall survival with Vectibix in metastatic colorectal cancer, after past studies have shown modest improvements in survival with Vectibix even in patients with wild-type KRAS tumors. Many panelists felt that Amgen has received too many chances already to show Vectibix has a meaningful impact on survival after having garnered accelerated approval for Vectibix four years ago. However, Amgen remained confident during the meeting that it would be able to show a survival advantage for Vectibix.
It is not clear how FDA would view Vectibix's accelerated approval status if the company is unable to show that the drug has a meaningful impact on patient survival. However, in recent years, the agency has placed increasing pressure on sponsors to complete post-marketing requirements after receiving accelerated approval. Furthermore, in oncology, the FDA is becoming stricter in demanding that investigational drugs show a survival advantage in post-marketing studies over the standard of care.
Commitment Issues
The FDA convened ODAC this week to discuss potential improvements the agency can make to its accelerated approval protocol, through which the agency expedites the market availability of drugs for life-threatening illnesses and for diseases with limited treatment options.
Accelerated approval is granted based on early signals on surrogate endpoints that a drug might have clinical benefit. However, sponsors are required to garner full approval for such drugs by completing post-marketing studies that show the treatment really does improve patient outcomes. Furthermore, under the FDA Amendments Act of 2007, sponsors can face monetary penalties for failing to complete post-marketing requirements with "due diligence."
The independent advisory committee heard from Amgen and four other sponsors that had outstanding post-marketing commitments related to accelerated approval NDAs. In the context of these applications, the FDA and ODAC members discussed whether the agency needs to change the clinical evidence requirements for drugs granted accelerated approval to ensure that sponsors were completing post-marketing studies in a timely fashion.
The discussion around Amgen's experience with Vectibix highlighted the challenges for a sponsor when new biomarker data emerges after accelerated approval is granted. "Biomarkers represent a regulatory challenge when there is not an FDA-approved diagnostic kit, as has been the case for the detection of KRAS mutations," Paul Eisenberg, VP of regulatory and safety at Amgen, said at the ODAC meeting. Knowledge of KRAS mutations as a negative predictive biomarker "impacted [Vectibix's] development program, since approval based on KRAS status … requires a combination drug and diagnostic approval pathway."
Currently, an FDA-approved KRAS test is still not available in the US market, though Amgen has a deal with Qiagen to develop a predictive test that is currently under review by the agency. Furthermore, the FDA is in the midst of ironing out regulatory requirements for drug/diagnostic combination products, such as Vectibix and its companion KRAS test. Therefore, sponsors developing such Rx/Dx products face significant regulatory uncertainty.
Although Amgen has fulfilled most of its post-marketing studies for accelerated approval of Vectibix, the company has been criticized by some for waiting too long to incorporate biomarker data in the development program for the drug, and choosing instead to pursue the largest patient population before embarking on a PGx study that would limit Vectibix's market share.
[ pagebreak ]
At a conference on personalized medicine in 2009, leading researchers and healthcare industry stakeholders questioned whether Amgen knew, well in advance of a negative regulatory decision on Vectibix by European regulators in May 2007, that patients with mutated KRAS genes would not respond to Vectibix. The fact that Amgen resubmitted its application to the European Medicines Agency a month later with KRAS-related drug response data led some at the conference to assert that Amgen initially pursued the largest indication possible for Vectibix with US and European regulators, knowing full well that a significant proportion of colorectal cancer patients would not respond (PGx Reporter 12/02/09).
In its defense, Amgen has previously said that while it had early hints about the association between KRAS mutations and Vectibix response, that data weren't mature enough to pursue changes to the drug's label. Eisenberg said at this week's advisory committee meeting that after its experience with Vectibix, the company has instated a policy of collecting patient samples and looking for biomarker targets early in the development program for all its drug candidates. Industry players developing Rx/Dx products, as well as the FDA, have recognized that sponsors need to plan ahead and incorporate biomarker analysis early in drug development programs.
Richard Pazdur, director of FDA's Office of Oncology Drug Products, explained that the purpose of the ODAC meeting was to figure out the types of conversations sponsors and the agency needed to have early on for drugs receiving accelerated approval. He noted that while the American public can understand why a drug with safety issues has been removed from the market, people have a harder time grasping the more nuanced reasons for granting accelerated approval for a drug and then revoking the approval later due to follow-on studies showing the drug lacks efficacy.
A recent example in this regard is Genentech's Avastin, which was granted accelerated approval in 2008 as a treatment for metastatic breast cancer. After confirmatory studies showed limited efficacy of the drug in this population, the FDA moved to take the drug the market — a decision that is highly controversial among patient groups and is being appealed by Genentech. Although Genentech has PGx data on Avastin from studies conducted by cooperative groups, the company only recently proposed conducting biomarker analysis to identify best responders, after FDA has threatened to revoke the drug's approval altogether (PGx Reporter 01/26/11).
Disruptive Biomarker
Amgen received accelerated approval for Vectibix from the FDA in September 2006, based on data from Study 20020408, a Phase III randomized-controlled trial in chemorefractory metastatic colorectal cancer patients comparing Vectibix plus best supportive care to best supportive care alone. In study '408, patients on the Vectibix arm had a statistically significant improvement in progression-free survival compared to those receiving just best supportive care. Although there was a 10 percent objective response rate among Vectibix-treated patients, increased overall survival was not demonstrated in study '408.
Around the time that Vectibix received accelerated approval, Amgen was also working on two Phase III studies: Study 2005181 comparing Vectibix and FOLFIRI versus FOLFIRI alone in the second-line setting; and Study 20050203 looking at Vectibix and FOLFOX compared to FOLFOX alone in the first-line setting. Study '181 had co-primary endpoints of progression-free survival and overall survival, while study '203 looked at progression-free survival as the primary endpoint.
The protocols for all these trials had to be amended, however, when in 2007 data emerged that KRAS-mutated colorectal cancer tumors did not respond to anti-EGFR monoclonal antibodies. At this point, Amgen amended the statistical analysis plan for study '408 to test for progression-free survival and overall survival in patients with tumors with wild-type KRAS.
After study '408 showed that Vectibix's efficacy was limited to patients who had wild-type KRAS tumors, the FDA made a class labeling change to recommend that EGFR-inhibiting monoclonal antibodies, Vectibix and Erbitux, should not be given to patients with KRAS mutations in codons 12 or 13.
[ pagebreak ]
The biomarker data also required Amgen to change the statistical analysis plan for studies '181 and '203. In '181, although patients with wild-type KRAS tumors saw statistically significant improvements in progression-free survival, overall survival wasn't statistically significant. Meanwhile, Study '203 showed similar results with regard to progression-free survival and overall survival in wild-type patients, but patients with mutated KRAS tumors fared worse on the Vectibix plus FOLFOX arm than with FOLFOX alone.
Since Amgen was unable to show a statistically significant improvement in overall survival with Vectibix, the FDA told the company that study '181 didn't fulfill the requirements to warrant full approval. After discussions, Amgen and FDA have settled on conducting one more study in chemorefractory metastatic colorectal cancer patients with wild-type KRAS tumors to try to show that Vectibix meaningfully improves overall survival.
How Many Bites of the Apple?
At the meeting, ODAC members debated whether Amgen was unnecessarily conducting a study to show improvements in overall survival with Vectibix.
"One of the purposes of doing these trials … is to look at earlier-stage disease. It's unlikely that these trials are going to show more than a few months of improvement in survival. And for the average patient this has very little impact," Patrick Loehrer, ODAC member and interim director of the cancer center at Indiana University, said at the meeting.
In KRAS wild-type patients, Study '203 in the first-line setting showed an overall survival of four months with Vectibix treatment, and in the second-line '181 trial patients on the Vectibix arm survived about two months longer than with just FOLFIRI.
Since several studies have shown that Vectibix has limited impact on overall survival, even in a molecularly defined subpopulation, committee members wondered why FDA was giving Amgen yet another shot. FDA's Pazdur agreed that this was an important point for consideration, but did not offer any definitive statements about what would happen if Amgen yet again failed to show a survival advantage for Vectibix. "How many bites of the apple do you get to take after you fail? … That's the real issue that needs to be addressed here," Pazdur said, adding that review of Vectibix's development program was ongoing.
Amgen appears confident that it can show a "meaningful survival advantage" for Vectibix. The company believes that survival data for Vectibix was confounded in the '408 study since investigators in that trial allowed patients whose cancer progressed while on the chemotherapy arm to cross over to Erbitux.
The company cited a trial involving Erbitux, a drug in the same class as Vectibix, which "demonstrated a survival advantage compared to supportive care alone in chemorefractory mCRC in patients with KRAS wild-type tumors," and that study design "did not include crossover from the control arm post-progression."
Taking cues from the above study, the Vectibix confirmatory study will prospectively select patients with KRAS wild-type tumors, and will not include a crossover to EGFR monoclonal antibody treatments if patients' disease progresses. Additionally, Amgen plans to conduct the trial in sites outside the US where anti-EGFR treatments are not the standard of care for metastatic colorectal cancer.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.