Yervoy Survival Data From First-Line Study Sealed The Deal For FDA Approval

FDA’s approval of Bristol-Myers Squibb’s metastatic melanoma immunotherapy was granted only after looking outside the BLA for assurance that the biologic’s effect was robust and validated.

FDA’s approval of Bristol-Myers Squibb Co.'s metastatic melanoma immunotherapy Yervoy (ipilimumab) was granted only after looking outside the BLA for assurance that the biologic’s effect was robust and validated.

Faced with an application for a second-line indication and a single study showing a survival benefit, but with only modest effects on secondary endpoints, the agency insisted on preliminary overall survival data from an ongoing first-line study to assure itself that approval was warranted.

The first-line study ended up addressing two FDA concerns about the original filing: it remedied problematic design issues inherent with the second-line trial, and it replicated ipilimumab’s effects on overall survival, thereby tipping the scale in favor of approval.

Although BMS requested a second-line indication, “there is no question that ipilimumab will de facto be used in the first-line setting,” FDA’s medical reviewer said.

The Yervoy experience shows the value to oncology drug sponsors of having a second study running at the time of BLA submission. Not only did the first-line trial give FDA confidence in ipilimumab’s effects, it also bought the sponsor a far broader indication than originally anticipated.

New Generation Of Melanoma Treatments

Ipilimumab is a first-in-class cytotoxic lymphocyte antigen-4 inhibitor that is believed to work by permitting development of an immune response to tumor antigens. BMS gained full rights to the drug through its $2.4 billion acquisition of development partner Medarex (Also see "What Medarex Has To Offer At $2.4 Billion" - Pink Sheet, 3 Aug, 2009.).

FDA approved ipilimumab on March 25 for the treatment of unresectable or metastatic melanoma, marking the first therapy for this type of cancer approved in more than a dozen years. Ipilimumab was also the first metastatic melanoma treatment cleared with a proven survival benefit and, as such, represents a major change in both the standard of care and the drug development landscape for a disease with few treatment options (Also see "Yervoy Approval Has Shifted Development Landscape For Melanoma" - Pink Sheet, 4 Apr, 2011.).

Yervoy’s approval was followed a mere five months later by FDA’s speedy clearance of another metastatic melanoma treatment, Roche/Genentech’s Zelboraf (vemurafenib). Unlike Yervoy, Zelboraf’s efficacy is pegged to the presence of a specific genetic mutation (Also see "Zelboraf Approval Hastened By FDA Officials Impressed With Early Efficacy" - Pink Sheet, 22 Aug, 2011.).

While the end result was all that BMS could have hoped for, FDA review documents show Yervoy’s development path took some unexpected turns and the BLA experienced a rocky review due to poor quality safety data (Also see "Poor Quality Data On Immune-Related Adverse Events Slowed Yervoy Review" - Pink Sheet, 1 Oct, 2011.).

First Came A Failed Trial …

Review documents indicate that BMS originally hoped to seek accelerated approval of ipilimumab monotherapy based on Study 184008 conducted under a Center for Drug Evaluation and Research IND. This was an open-label, single-arm trial of ipilimumab (10 mg/kg) in 155 subjects with previously treated advanced melanoma.

Study 008 assessed objective response rate as a surrogate endpoint. However, the results failed to meet the threshold level for efficacy specified in the statistical analysis plan. Consequently, during an April 2008 pre-BLA meeting, the agency recommended BMS await overall survival data from an ongoing Phase III study (CA 184024) to support a BLA filing.

Study 024 originally was intended to serve as the confirmatory trial for Study 008. The study enrolled more than 500 patients receiving initial treatment for metastatic or unresectable melanoma; subjects were randomized to dacarbazine with or without ipilimumab 10 mg/kg. Although the original primary endpoint was progression-free survival, BMS subsequently amended the protocol to make overall survival the key outcome.

While Study 024 was under way, BMS met with FDA to discuss whether the survival results from another study, MDX010-20, conducted under a Center for Biologics Evaluation and Research IND, could serve as the sole trial supporting ipilimumab’s approval.

Study 20 was a randomized, double-blind trial conducted in HLA-A*0201-positive subjects with Stage III or IV melanoma who had relapsed or progressed after previous therapy; 676 patients were randomized 3:1:1 to receive either ipilimumab 3mg/kg plus an experimental gp100 melanoma peptide vaccine, ipilimumab monotherapy or the gp100 vaccine alone.

While the study was still blinded, the primary endpoint was changed from best overall response rate to overall survival. Initial data presented to FDA showed a significant 3.7-month improvement in median overall survival for ipilimumab alone compared with gp100 alone, with a hazard ratio of 0.66 (a 34% risk reduction).

The ipilimumab/gp100 arm also fared significantly better than the vaccine group alone, demonstrating a 3.51-month improvement in median survival with a hazard ratio of 0.68 (a 32% risk reduction).

… Then One With Questionable Results …

FDA agreed the Study 20 survival results were of interest, but decided survival data from Study 24 would be needed to provide additional confirmatory evidence of efficacy, particularly given its concerns about Study 20’s design and conduct.

In a Feb. 25 review, Division of Biologic Oncology Products Medical Officer Kaushikkumar Shastri said there were concerns as to whether the overall survival results in Study 20 were driven by a poor outcome in the gp100 control arm.

Another issue was the trial’s population, which was limited to patients with HLA-A2+ phenotype. “This was done … in order to adequately assess response to the vaccine component, which required HLA-A2 phenotype to ensure successful presentation of the peptide vaccine,” Shastri explained. “However, there is a possibility that the HLA selection contributed to the positive results seen in the study.”

In a Feb. 25 memo, statistical reviewer Yuan-Li Shen cited concerns about the reliability and robustness of ipilimumab’s treatment effect based on a single small study. There also were numerous unplanned changes to the Study 20 protocol that may have introduced bias.

Furthermore, the effects on secondary endpoints such as best overall response rates and progression-free survival were too modest to support the overall survival results, FDA reviewers said. The best overall response rates in Study 20 were 5.7% for the combination arm, 10.9% for ipilimumab monotherapy and 1.5% for gp100 monotherapy. Investigator-assessed PFS was similar in all three treatment arms, ranging from a median of 2.76 months in the gp100 monotherapy and ipilimumab/gp100 arms to 2.86 months in the ipilimumab monotherapy arm.

… But First-Line Data Seal The Deal

Since Study 24 was ongoing at the time of BLA submission, the parties agreed that BMS would provide top-line data to FDA in such a way that all but one employee at the company remained blinded to the results (“Protecting The Blind: FDA, BMS Took Pains To Ensure Integrity Of Ongoing Yervoy Study,” Pharmaceutical Approvals Monthly, October 2011).

Although none of the FDA reviews go into extensive details about the Study 24 results, it is clear that the survival data in the first-line setting sealed the deal for the review staff.

The high-level results provided to FDA were based on approximately 97% of the planned number of events for the final survival analysis. “The results … show an improvement in overall survival [HR 0.85 (95% CI: 0.76, 0.93)], with a nominal p-value of 0.001, stratified log-rank test,” Division of Biologic Oncology Products Director Patricia Keegan wrote in her March 12 review. “The Kaplan-Meier curves for overall survival in this trial are similar to those in MDX010-020 in showing a relatively early separation of curves and that is maintained throughout later timepoints.”

The Study 24 results not only nailed down the approval, they also helped negate the need for an Oncologic Drugs Advisory Committee review of the BLA. A planned ODAC meeting on the ipilimumab application was canceled on two separate occasions.

ODAC’s advice ultimately became unnecessary, Keegan explained, as demonstration of a survival benefit in both trials “provided sufficient evidence of clinical benefit, particularly in this malignancy where there is currently no highly effective treatment. Evidence of improved survival also provides assurance that the benefits generally outweigh the risks.”

Comfortable With A Broad Indication

The robust Study 24 results, especially in a therapeutic space with few treatments, moved FDA to grant a broad indication spanning genetic profiles and lines of treatment. Unlike Study 20, Study 24’s population was not restricted according to HLA phenotype. Consequently, FDA did not limit the indication to only those patients with a certain genetic marker.

FDA’s decision not to limit the indication to second-line or later use was also guided by practical considerations. “The application originally specified an indication for ipilimumab for the treatment of advanced melanoma in patients who have received prior chemotherapy,” medical reviewer Shastri said. “However, since there is no effective first-line therapy and none has shown survival advantage, there is no question that ipilimumab will de facto be used in the first-line setting by the oncology community.

“Therefore, following the receipt of top-line survival curves for the nearly completed first-line study of ipilimumab, the review division and the office is comfortable in giving ipilimumab an indication for advanced melanoma with or without prior chemotherapy.”

However, the review division was not comfortable allowing labeling to discuss use of ipilimumab for retreatment. In Study 20, 40 subjects (5.9%) received retreatment following progression: 29 in the ipilimumab/gp100 treatment arm, nine in the ipilimumab monotherapy group and two in the gp100 arm. “The safety and efficacy experience with retreatment was thus very limited, hence it cannot be recommended as a treatment strategy,” Shastri said.

Despite its endorsement of ipilimumab’s efficacy, the review division insisted that BMS do more work on dose optimization, as the two studies show survival advantages at two doses. The approval letter includes a post-marketing requirement for a trial comparing the efficacy and safety of the 3 mg/kg and 10 mg/kg doses in patients with unresectable Stage III or IV disease. Patient accrual is to begin in March 2012, with a final study report due in December 2017.

[Editor’s note: This story appears courtesy of Elsevier Business Intelligence’s Pharmaceutical Approvals Monthly. For more information, please contact customer care at 800.332.2181 or [email protected], or sign up for a free trial at http://www.elsevierbi.com/publications/pharmaceutical-approvals-monthly/free-trial .]