In FDA's review of Lilly/Daiichi Sankyo's antiplatelet agent Effient (prasugrel) - a drug with broad potential use - it is notable that the approval was delayed not by unusual clinical questions, but rather by the agency's efforts to be exceedingly thorough and scrupulous in the review.

FDA review documents suggest that review management issues contributed more to the lengthy and overdue review of Effient than any unresolved clinical debates did.

"The issues were pretty clear early enough to have allowed us to meet the original PDUFA goal," Cardiovascular and Renal Products Division Director Norman Stockbridge declared in an April 25, 2009, decision memo.

It took FDA another two and a half months, and more than a year after the original June 26, 2008 user fee goal for the priority review, to finally approve Effient on July 10, 2009.

The three major issues in the review - weighing the risk of bleeding with the benefit of averted cardiovascular events; determining the tumor promotion potential of prasugrel given an imbalance of neoplasm findings compared with clopidogrel; and an unprecedented CMC issue, a conversion from prasugrel salt to base discovered late in the pivotal TRITON-TIMI trial - were well known and had been extensively discussed as the review proceeded.

"This has not been a well-managed review process," Stockbridge stated. "What has been missing is a clear means or will to declare an end to discussions and to allow the regulatory process to complete. No one associated with this review should feel good about this."

An accident of timing put Effient under review just as FDA was letting up on pressure to complete reviews by the user fee goal date. Office of New Drugs Director John Jenkins determined that meeting PDUFA goals would not be a priority while FDA integrated new responsibilities and authorities under the FDA Amendments Act. (¹ Also see "The New User Fee Rules: FDA Sacrifices Timeliness, Tries to Save Predictability" - Pink Sheet, 1 March, 2008..)

Yet Effient stood out for its prolonged review even among the group of drugs where FDA decided the user fee goals would not be met. Most missed goal products were approved shortly after the due date. Effient's user fee date was extended by the submission of a major amendment with additional sponsor analyses of significant questions, but it still did not receive its first action letter, the approval, until ten months after the revised deadline.

Jenkins, an FDA official usually above the normal review process, did provide the sponsor with some explanation for the delay - which goes to show the extreme caution the agency was taking with such a high-profile product.

"There are many uncertainties and perspectives," Jenkins told the sponsors during a December 2008 teleconference, and FDA "needs to be comfortable when taking a regulatory action," according to FDA's minutes. "The agency will need to be able to explain and defend whatever action it takes."

All Eyes On FDA

The review of Effient was unusually extensive, but then, the drug has been touted as potentially revolutionizing a massive market.

And FDA had to have been conscious that Effient was likely only the first in a series of coming NDAs for monster anticoagulant products, such as AstraZeneca's Brilinta (ticagrelor). Brilinta, which could be the first reversible antiplatelet agent, is on track for a late-2009 NDA (² Also see "AstraZeneca's Ticagrelor Brilliant At ECS Congress; Will It Outshine Effient?" - Pink Sheet, 7 September, 2009.).

With a high-pressure atmosphere, it is not surprising the reviewing division at FDA sought out additional input. In addition to lateral advice in the form of consults from specialists - five consults are incorporated into the review, on the high end for new drug reviews - the division brought vexing issues to the center level in two regulatory briefings for high-ranking officials, including CDER Director Janet Woodcock and OND's Jenkins. Regulatory briefings, sometimes characterized as "internal advisory committees," allow review staff to gain input on difficult issues from center leadership. (³ Also see "Regulatory Briefings: FDA's "Internal Advisory Committees" " - Pink Sheet, 1 November, 2008..)

The review already had high-level input. Stockbridge's deputy director in the Division of Cardiovascular and Renal Products, Ellis Unger, became acting deputy director of ODE I during the review (a position he now holds on a permanent basis).

Unger played a key role in the Effient review, authoring both a secondary review and a set of memos distilling the cancer risk, bleeding risk and the CMC form conversion issues. [Editor's note: Our coverage of the Effient review, including the risk of excess bleeding and the form conversion issue, will continue in the next issue of Pharmaceutical Approvals Monthly.]

Those were all complex issues that were significant matters for the review, and added up to a significant delay, but scrutiny of the review documents (and Stockbridge's comments) indicates that FDA's attempts to be vigilant may have devolved into a circular rehashing of the issues.

The potential risk of malignancies, a late-breaking concern, is a prime example of how review management was a factor, and highlights tactics that played out in the handling of the review - including multiple consults and additional rounds of analyses.

Metastasis Of Cancer Risk Review Reflects Desire To Be Meticulous

Cancer risk was a particularly ripe target for internal FDA consultation. Both the FDA review team and the sponsor recognized that the prasugrel group in the pivotal TRITON trial (also referred to as TAAL by the review staff) had more malignancies than the group receiving the active comparator, Bristol-Myers Squibb/Sanofi-Aventis' Plavix (clopidogrel).

Despite that baseline agreement, "there have been countless internal discussions, blinded analyses by staff of new (or possibly new) tumors, consideration of whether our primary interest was in 'new' or 'new and worse' tumors, and statistical analyses," Temple said in his July 10, 2009, approval memo.

Facing the imbalance in malignancies, the Cardio-Renal division looked for expertise both internally and externally. The review team first sought a consult from the Division of Drug Oncology Products, which in turn urged a review by the Office of Surveillance and Epidemiology.

The OSE Division of Epidemiology consulting review on the cancer risk was distinct from OSE's intense involvement in the development of the REMS for prasugrel. Unger remarked upon the "extensive internal discussions and consultation with the Office of Surveillance and Epidemiology" over the REMS in his Jan. 9, 2009, review. Effient was approved with a moderate REMS (⁴ Also see "Effient REMS: FDA Role In Intro Grows; No Safe Use Restrictions" - Pink Sheet, 20 July, 2009.).

OSE Urges Advisory Committee Review

It was the OSE consult on the cancer risk that may have led to the advisory committee meeting on Effient. Epidemology Team Leader Allen Brinker recommended taking the Effient debate public, counter to the Cardio-Renal division's decision not to seek a panel review. He deemed it "prudent to present" the Effient NDA to FDA's Cardiovascular and Renal Drugs Advisory Committee.

Many important outstanding questions "should be vetted and discussed before a public advisory committee before another prasugrel trial, let alone before the general approval and marketing of prasugrel," Brinker urged in a June 13, 2008 memo.

Somewhat ironically, FDA had initially decided against an advisory committee in order to avoid delay in bringing an important public health advance to market quickly.

"In light of what appeared to be robust efficacy findings, the Division, with concurrence of the Office, decided initially that the application should forego a public Advisory Committee meeting," Unger recalled. "Given that prasugrel appeared to be superior to established treatment for the prevention of non-fatal MI, this approach was planned in the interest of public health, so that regulatory action would not be unnecessarily delayed."

The detection of "unanticipated issues [that] came to light during the review process" persuaded the review team that an advisory committee would be warranted: the imbalance in neoplasms suggesting cancer risk and the form conversion from salt to base. "Other individuals thought that a public discussion of the bleeding risk would be of value," Unger added.

It seems that officials above the division level also wanted the public vetting. Unger's Jan. 9, 2009, memo reports that, "Ultimately, the Office reached the conclusion that a public presentation of these issues to the Cardio-Renal Drugs Advisory Committee would be advisable."

Unger observed that the multiple issues of bleeding risk, excess malignancies and form conversion had generated "considerable discussion" across the staff - including the chemistry, pre-clinical pharmacology-toxicology, clinical pharmacology, and clinical review staff within the Division, as well as staff within the Division of Drug Oncology Products, Office of Surveillance and Epidemiology, and Office of Drug Evaluation I. That not only illustrates FDA's drive for the most complete assessment of risk before releasing a high-profile potential blockbuster on the market, it also points to the desire for a transparent process to avoid criticism later on.

The Division's Point Man On Cancer Findings

Inside the Cardio-Renal division, the prasugrel NDA team drafted medical team leader Tom Marciniak to contribute a "special secondary review" looking at cancer adverse events and risk/benefit. Marciniak, who took a very personal approach to the task, said he "initiated the analyses because of my assignment as the clinical reviewer for the prasugrel IND, a professional interest in exploring cancer rates in large outcomes trials, and the suggestive results (in my interpretation) of the mouse carcinogenicity study."

"Because my preliminary analysis raised the issue of increased cancer rates with prasugrel in a large outcome study, the Cross Discipline Team Leader for this submission requested that I complete and formally submit my analyses," Marciniak explains in his May 6, 2009, review.

Unger noted that Marciniak held a "minority view" in contrast to the rest of the review staff regarding the carcinogenicity studies; the majority did not regard them to be positive for a cancer signal for prasugrel.

First Hurdle For Cancer Assessment: Data Quality

"This is a very complex submission," Marciniak observed. Moreover, "the quality of the data in TAAL is substantially worse than the quality of the data in other large outcome trials that I have reviewed, e.g. LIFE, RALES, EPHESUS, CAPRIE, CREDO, CURE, CHARISMA and COMMIT," Marciniak said. The "C" trials are of particular interest as clopidogrel outcome trials.

The problems with the pivotal TAAL trial combined poor data quality and collection parameters. "TAAL was not pre-specified to examine cancer rates, although cancer events are routinely captured in most CV trials," Marciniak noted.

While other reviewers determined that the lack of cancer histories resulted in a data quality issue that was "insurmountable," Marciniak concluded that "does not make the TAAL cancer results uninterpretable." Interpretation, however, required significant evaluation of individual case report forms.

"Because of disparity between the accounting of the cases by Dr. Marciniak and the applicant, much additional attention was given to obtaining agreement on the actual numbers of new, non-benign neoplasms," Unger said. That additional attention came from high-ranking drug review staff, and again shows the willingness of top FDAers to get involved: Unger notes that he, Marciniak, Division Director Stockbridge and Office Director Temple "blindly adjudicated a subset of the cases." Unger noted that Marciniak continued to identify additional cases for re-classification.

FDA Ascertains No Ascertainment Bias

Lilly was actively, even aggressively, involved in the Effient review, from the unusually extensive rounds of meetings during the IND phase to frequent interchanges throughout the NDA review. The cancer situation showcases that activist approach.

Lilly energetically promoted its own interpretation of the cancer risk data, ascribing the TAAL neoplasm findings to "ascertainment bias." As Unger relates in a July 7, 2009, review memo on the cancer risk, Lilly focused on the 30-40 percent increase in bleeding events with prasugrel relative to clopidogrel. Lilly posited that "a disproportionately greater frequency of bleeding events in the prasugrel group would lead to a disproportionately greater number of patient evaluations, which would uncover disproportionately more cancer cases," Unger explained.

"Although the theory seemed plausible on its face, the Division undertook its own analysis of the cases," Unger reported. "The between-group difference in neoplasms largely persisted."

Lilly pursued the issue. The company submitted additional TAAL analyses and information from the ongoing endeavor to identify cancer cases in the white noise of sometimes poorly-completed case report forms.

However, even Lilly's substantial submission of clinical information and analyses in May 2008 failed to convince the Cardio-Renal division that the cancer signal was an artifact of ascertainment bias.

FDA followed a different intellectual path, but ultimately came down close to Lilly on how strongly to react to the tumor promotion concern - opting for restraint while awaiting more study. "No one believes a definitive answer is possible without further data," Temple emphasized.

Temple listed other reasons to be skeptical of a tumor promotion effect of prasugrel. "Dr. Unger is skeptical about the reality of the increased rate of new tumors," he reported, "given the benign animal data, including the recent tumor progression studies, the absence of examples of such tumor promotion or any mechanistic explanation, and the inherent multiplicity of our safety analyses." Temple added another reservation: "the marginal statistical support for the increase."

The company's campaign didn't fall on deaf ears: Unger noted Lilly's ascertainment bias argument "does provide some measure of reassurance."

Unger further pointed to the disjunction in mortality due to cancer, with 27 prasugrel and 19 clopidogrel deaths. An "imbalance in cancer deaths is concerning, because mortality would not be expected to be affected by ascertainment bias," he commented.

"Beyond a mere association between prasugrel and excess cancers, therefore, biological plausibility, exposure response, and other factors are helpful to support causality, and these factors seem to be missing here," Unger concluded (see sidebar on ⁵ Also see "Imperfect Information: Since Tumor Promotion By Prasugrel Was Unprovable, FDA Asked If It Was Biologically Plausible" - Pink Sheet, 1 October, 2009. ).

Do Tumor Concerns Warrant Limiting Duration?

Although Temple made clear "no one in the Division of Cardiovascular and Renal Products believes the cancer findings should block approval," Unger's memo indicates that there were "differing opinions in the Division as to how labeling should be handled."

"There are some who argue that if there is a risk of tumor stimulation, it should be related to exposure," Unger reported. "These individuals advocate placing a limit on the duration of prasugrel use to perhaps 30 days, with patients switching to clopidogrel at that point." Chief among "these individuals" was the Cardio-Renal division's Marciniak.

Despite difficulties with the trial data, Marciniak felt that cancer promotion data justified restrictions on prasugrel to limit patient exposure to the drug by switching to clopidogrel after a month.

"The evidence for a problem is far stronger in TAAL than it was at NDA submission times for the recent withdrawals from market," Marciniak declared.

"I can argue that the short term benefit justifies immediate approval, although only for short-term use," Marciniak stated. "But I can also argue that approval should be delayed until the planned trial in medically managed [acute coronary syndromes] addresses the cancer promotion issue." That data is coming from Lilly's TABY (or TRILOGY) trial, due in January 2013.

The primary clinical reviewer, Karen Hicks, initially also supported limitations on the duration of prasugrel use to as little as one week. However, in a July 8, 2009, review memo, Hicks "reached the opposite conclusion, and strongly urges long-term treatment, largely because her detailed analyses showed that prasugrel gives a substantial … reduction in stent thrombosis, a potentially fatal event that is also a late event," Temple reported.

"Any proposed duration of treatment is necessarily arbitrary," according to the anti-limited use position, Unger said. "Switching involves logistical issues" that could lead some patients to "simply discontinue their thienopyridine, which could lead to stent thrombosis." Additionally, "the strategy of switching from prasugrel to clopidogrel has not been tested," he said.

"For the majority of the review staff who believe more strongly that the imbalance is spurious, the exposure issue is moot, and they would not place any limitation on duration of use," Unger said, adding: "I agree with the majority view on this issue."

Temple also agreed; approved Effient labeling does not place limits on duration of use.

Where In Labeling Should Cancer Risk Go?

"Risk of cancer is always of great interest … and cannot be ignored. A precaution seems appropriate for labeling at this time, although others have argued for a warning or boxed warning," Unger noted.

The advisory committee panelists were even more skeptical of the cancer promotion effect, urging placement of the cancer risk information in the adverse reactions section of labeling instead of warnings/precautions, where draft labeling had placed the language.

The committee raised some eyebrows by downgrading FDA's draft labeling. FDA had presented three options for addressing the cancer concern in labeling: in warnings/precautions, in a black box, or the most strict, requiring limited use (⁶ Also see "The Return Of The Blockbuster? Stage Is Set For Prasugrel Approval" - Pink Sheet, 9 February, 2009.).

Stockbridge agreed that cancer risk "merits minor mention in the label's Adverse Reactions section only." Approved labeling places the cancer data with adverse reactions, boiling down the lengthy debate into a brief paragraph summarizing that "newly diagnosed malignancies were reported in 1.6 percent and 1.2 percent of patients treated with prasugrel and clopidogrel, respectively," but "it is unclear if these observations are causally related or are random occurrences."

Advisory Committee Offered Valuable Voices

The Feb. 3, 2009, advisory committee, with a unanimous endorsement of Effient approval, provided a valuable counterweight to FDA's skittishness about resolving the most contentious issues in the NDA. And the committee's public and confident endorsement did not fulfill FDA's concerns, early in the review, that an advisory panel might delay approval of a clearly efficacious drug.

On the other hand, the run-up to the meeting produced an incident that unleashed a wave of controversy about FDA's advisory committee management. Just days before the meeting, Lilly, continuing its aggressive pursuit of Effient approval, objected to potential "intellectual bias" on the part of committee member Sanjay Kaul (Cedars-Sinai Medical Center). FDA hurriedly "disinvited" the cardiologist, who had presented critical analyses of the drug at a medical meeting. (⁷ Also see "The Prasugrel Backlash: Another Black Eye for FDA's Advisory Committee Process" - Pink Sheet, 1 April, 2009..)

Lilly's action proved counterproductive. Kaul was eventually invited to present his analyses to OND topsiders, who heard nothing that the review team had not already considered (see sidebar, "⁸ Also see "Advisory Cmte. Disinvitee Presented Prasugrel Concerns To Top CDER Staff" - Pink Sheet, 1 October, 2009.").

However, the perception that FDA had tried to "stack the deck" in favor of prasugrel likely contributed to the internal perception that the review process must be unimpeachably thorough. Indeed, despite the committee's unanimous endorsement, FDA took five more months to reach its decision.

Postmarketing Requirement Vs. Commitment

In assessing Effient's potential risk of malignancies - a significant issue for a primary care drug with such wide potential use - FDA considered all of its options for collecting better data, including its FDAAA-granted authority to mandate postmarketing studies.

"There is discussion of making collection of cancer data a PMR [post-marketing requirement] for the ongoing outcome study of prasugrel in medically managed ACS," the 10,300 patient TABY (also known as TRILOGY) trial, Stockbridge said.

Unger was on board with the PMR approach. He agreed that the ongoing TABY trial could be used to assess prasugrel's role in stimulating cancer. The trial could specifically be used to screen for cancer, identify pre-existing tumors and track definitions and classifications of tumors.

"The sponsor has incorporated reasonable data collection," Stockbridge observed, "and I do not see how making this a PMR is useful, since we cannot expect them to continue the study for this purpose if there is any reason to abort it."

But Stockbridge was overruled: cancer data from TRILOGY/TABY is a post-marketing requirement (⁹ Also see "Effient Approval Shows Mark Of FDA Controls" - Pink Sheet, 15 July, 2009.).

- Bridget Silverman ( ¹⁰ b.silverman@elsevier.com )

Effient Review Delayed More By Management Issues Than Clinical Disputes

More from Archive

More from Pink Sheet