Shepherding the orphan Huntington's chorea drug Xenazine through a three-year review, top-level FDA reviewers invoked the concept of "confirmatory evidence" to circumvent the agency's preference for two randomized clinical trials to support new drug approval.

At the first two levels of review, "not approvable" actions were recommended by clinical efficacy reviewer Carole Davis and Neurology Team Leader John Feeney, because the NDA evidence did not fit the statutory parameters, delineated in a May 1998 guidance, for "substantial evidence of effectiveness."

When Xenazine (tetrabenazine) came before Division of Neurology Products Director Russell Katz, the discussion turned to applying the guidance's alternative provision for a randomized clinical trial plus confirmatory evidence.

"Ordinarily, it should be noted, a 'failed' second study should not be considered to 'confirm' another, 'positive' study," Katz stressed in a March 18, 2008, memo. For Xenazine, one clinical trial found robust evidence of effectiveness, but the other, smaller trial was undermined by human error.

Upon its August 15, 2008 approval, Prestwick's Xenazine (tetrabenazine) became the first drug approved for Huntington's disease. The NDA drew from multiple sources. In addition to the two randomized clinical trials, data was submitted from the 20 years of use under an IND at Baylor; even earlier data was provided from Roche, which had marketed tetrabenazine in Europe as an antipsychotic under the brand Nitoman .

FDA's first "approvable" letter for Xenazine, dated March 24, 2006, is, at eight pages, unusually long - and goes into atypical detail to explain the agency's reasoning in accepting the efficacy data as adequate. The letter also raises safety issues. While the concerns about whether the data meets the efficacy standards were resolved in the first cycle, it took another two review cycles to work through the safety issues.

The letter, signed by Office of Drug Evaluation I Director Robert Temple, demonstrates that the upper echelons of the agency were more willing to apply the confirmatory evidence standard for Prestwick's orphan drug. "The findings establish the effectiveness of Xenazine as a treatment for the chorea of HD, under FDAMA's provision that substantial evidence can consist of the results of a single adequate and well-controlled investigation plus confirmatory evidence," the letter states.

"We believe that the statistically strong results of Study 004, its marked internal consistency, as well as the results of Study 005, provide the necessary confirmatory evidence required by this provision of the Act," the approvable letter continues.

Team Leader Feeney had ruled out using the 'one randomized trial plus confirmatory evidence' standard based on the guidance's description of what circumstances could invoke that exception to the two-trials rule.

Feeney noted that FDA's May 1998 guidance on clinical evidence of effectiveness "allows for approval with a single study plus confirmatory evidence." The confirmatory pathway "is generally intended for cases where there is an effect on mortality or irreversible morbidity," he said. "That is not the case here."

The review team leader raised the question of whether Xenazine is a case that warrants deviating from the "usual requirement" for two independent studies - noting, "HD is a tragic disease. There are no drugs." But there were additional challenges posed by the Xenazine data set.

Second Study Not A Strong Set Of Data

Only one trial - Study 004, the larger of the two with over 100 patients - was provided as evidence of effectiveness of Xenazine on management of chorea, one symptom of Huntington's disease. Fortunately for Prestwick, the results for the primary endpoint were robust. "The strength of evidence provided by Study 004, alone, seems to establish the efficacy of TBZ after three months of treatment," Feeney stated.

The other controlled trial in the NDA, Study 005, sailed on choppier waters - mostly because of poor execution. In his Aug. 15, 2008, decision memo, Temple referred to the "travails" of Study 005, "a small randomized withdrawal study in just 30 patients." Study 005 "did not meet the usual standard (p=0.05) for being a 'positive' study," Katz noted in his March 18, 2008, memo. "The sponsor suggested that this result was related to a study conduct issue, specifically that patients inappropriately had their morning dose of active drug withheld on the morning of Day 3."

Prestwick conducted post hoc analyses of the Study 005 data to redeem the utility of the study after discovering the protocol breach, which helped to persuade FDA to consider Study 005 as confirming Xenazine's efficacy. The results "were numerically very close to 004 in the 18 patients treated correctly, so that study provides some support for Study 004," Temple said.

The strength of the specific chorea results of Study 004 also weighed heavily in FDA's final assessment. In 004 "the effect on chorea seemed so robust," Katz's approval memo notes, with an "extraordinarily" low p-value and consistency across the centers (striking given how small the sites were) that was supported by the pattern of response in patients re-treated in later studies.

Chorea Benefit Clear, But Other Benefit Not Seen

However, compelling as the chorea data was, the lower level reviewers were troubled that the chorea reduction was not associated with any other patient benefit - and that contributed to their hesitations about not having two trials.

"If the sponsor could present evidence that the change in chorea seen in Study 004 resulted in some measurable benefit to patients (patient global or functional scale improvement), then I would argue to suspend the usual requirements and approve TBZ with the one study plus confirmatory evidence standard," Feeney said. "But that is not the case."

There is "no evidence of a functional correlate with the decreased chorea" in Study 004, he stated, and "there is no patient measure of satisfaction."

As the primary reviewer Davis observed, "lack of expected other effects accompanying a critical outcome can weaken the persuasiveness of a single trial." FDA had anticipated an effect on functional endpoints as well as chorea scores, she noted. Tetrabenazine "had been given a priority review status primarily based on the assumption that improvement of chorea control would result in improvements in gait safety and functional activities (activities of daily living)."

"The study was not able to show significant treatment benefit in these areas, and results generally trended toward the placebo group," Davis reported. "On cognitive assessment, the placebo group was favored at a statistically significant level."

Additionally, Feeney observed the trial raised questions of potential adverse effects on non-chorea elements of HD. "There is some evidence of a cognitive decrement" and "some evidence of functional decrement," he said. The safety profile only added to that, with links to depression, suicidality, Parkinsonism and worsened dysphagia , which increaseses the risk of aspiration pneumonia.

"Given these points, this does not seem to me to be the place to abandon the usual two-study standard," Feeney concluded.

As the primary reviewer, Davis - who had the least latitude in interpreting FDA's standards - recommended a "not approvable" action, though she indicated an approvable action would be appropriate, were the other data "allowed to contribute to the decision." Constrained by the strict interpretation of the two-trial requirement, she called for another confirmatory study. "The population of HD patients is large enough that enrollment for an additional study should not pose a serious obstacle for the sponsor."

"I think there is something valuable that can be gained by further study," Feeney added. "And, pending further study of TBZ, neuroleptics are available for the off-label treatment of chorea."

A new randomized trial could evaluate whether efficacy is maintained over time, and "if the efficacy is maintained, at what cost to the steadily declining cognitive, behavioral, and functional status of patients with Huntington's disease?" Feeney said.

Reinterpreting The Efficacy Standard

Katz considered the two "not approvable" recommendations from the efficacy review team in a March 16, 2006, memo, but, with the full authority of a division director, pressed ahead with the concept of "confirmatory evidence" to unlock the potential "approvability" of the drug based on the existing trials.

Katz acknowledged that the sponsor did not meet the "typical requirement" of evidence from two adequate and well-controlled trials. Further, he acknowledges that the Xenazine data do not necessarily seem to qualify for the statutory standard of substantial evidence from a single adequate and well-controlled trial plus confirmatory evidence.

"The agency guidance … suggests that such a standard should generally be applied when the drug has an effect on mortality or irreversible morbidity and when such a trial could not be practically or ethically repeated," Katz noted.

"I do not believe that these criteria have been met in this case," his memo states. "The effect on chorea is symptomatic, and there would seem to be no ethical impediment to performing another trial."

"However, the statute offers a standard that might be met by this data package," Katz states. "Many of the elements that the 'evidence' document suggests could serve to support the use of a single trial (putting aside … that the effect is not one on mortality or irreversible morbidity) are present here," he said, "including a very small p-value, equivalent effects in sub-groups of different disease severity, and numerical superiority of drug compared to placebo in 14/15 study sites."

Katz concluded those findings could be considered to constitute confirmatory evidence. "Although this language is not used in the 'evidence' document," he notes, "these findings are included in that document to serve exactly the same ends as confirmatory evidence; that is, to support the use of a single adequate and well-controlled trial as providing substantial evidence."

Other Review Cycles, Other Issues

Resolving the adequacy of the efficacy evidence allowed the review to proceed, but it did not clear the way for approval. The next two approvable letters - and an advisory committee meeting - focused more on issues of safety and net benefit of the drug. Even accepting the evidence for efficacy as sufficient, the original approvable letter notes that the NDA posed problems. "Despite the documented effect on chorea, there remain troubling questions about the utility and ultimate approvability of this application," it states.

"We were unsure that the potential benefit of tetrabenazine on chorea could be justified," Katz reported in a March 18, 2008, memo.

"The issue that has concerned FDA has been that, despite the clear effect of tetrabenazine on choreiform movements, many other pertinent outcomes were affected adversely," Temple said in his August 15, 2008, memo. He noted the drug's significant adverse events, notably parkinsonism, akathisia, depression and dysphagia, as well as [worsening of] other elements of the disease itself.

"Notably, functional assessment and cognitive assessment … were statistically worse," Temple noted. Behavior assessment, independence scales and functional capacity "were directionally, albeit not significantly, worse."

"This bidirectional set of effects raised the question of whether there was a 'net benefit,'" Temple said.

"We were concerned that if the drug actually caused deterioration in these [functional and cognitive] domains, it would be difficult for the practitioner to recognize these clinical changes as being drug-related, given that deterioration of function and cognition are symptoms of HD itself," Katz said.

Katz also emphasized "the clear drug-related increase in significant adverse events, including Parkinsonism, akathisia, depression, and dysphagia, the latter possibly being associated with aspiration pneumonia."

"Here, too, we were concerned that practitioners might to be able to identify some of these events as being drug related, again because several of these are symptoms of HD," Katz said. "In particular, if these events were drug related, but were not considered as such, it is possible that they could continue to increase in severity, perhaps becoming irreversible and resulting in significant clinical sequelae."

Those lingering safety concerns should be addressed by the Risk Evaluation and Mitigation Strategy developed by FDA and Prestwick. The REMS for Xenazine not only helped gain approval, it helped allay investor concerns as well, and contributed to a cascading series of deals that have brought the drug to Lundbeck. The REMS, which calls for extensive educational outreach, essentially codifies a marketing strategy, and - as the billion dollars worth of deals surrounding Xenazine show - raised expectations that the drug should become an orphan marketing success (¹ The RPM Report, February 2009).

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

FDA's First Step In Xenazine Review: Work Around Clinical Evidence Guidance

More from Archive

More from Pink Sheet