FDA Panel Will Weigh Northera’s Short-Term Benefit Vs. Lack Of Durable Effect, Safety Concerns

At its Feb. 23 review of Chelsea Therapeutics’ investigational orthostatic hypotension drug droxidopa, FDA’s Cardiovascular and Renal Drugs Advisory Committee will be asked to discuss sufficiency of the short-term efficacy data, clinical meaningfulness of the effect size and safety issues, including Japanese post-marketing reports of neuroleptic malignant syndrome.

Whether the short-term benefit seen with Chelsea Therapeutics International Ltd.’s Northera (droxidopa) on the cardinal symptom of neurogenic orthostatic hypotension outweighs the lack of a demonstrated durable effect and questions about long-term safety is the issue facing FDA’s Cardiovascular and Renal Drugs Advisory Committee.

At the panel’s Feb. 23 review of droxidopa, the agency will also seek the committee’s views on whether the effect size seen with droxidopa is clinically meaningful and the adequacy of the sponsor’s proposed post-marketing study to further characterize the drug’s safety profile.

Although the agency does not pose questions to the panel about the patient-reported outcomes instruments used in the droxidopa development program, review documents released ahead of the meeting challenge the validity of those tools and suggest that even if Northera were approved, the drug is not likely to receive a broad claim for symptom and functional improvement.

Droxidopa Seeks To Join Midodrine

Chelsea is seeking approval for the treatment of symptomatic NOH in patients with primary autonomic failure (Parkinson’s Disease, multiple system atrophy and pure autonomic failure), dopamine-beta-hydroxylase deficiency and non-diabetic autonomic neuropathy.

NOH is an orphan indication with few good therapeutic alternatives.

While some drugs are used off-label, midodrine (Shire PLC’s ProAmatine and generics) is the only currently approved treatment for orthostatic hypotension. It received accelerated approval in 1996 on the basis of a blood pressure endpoint – a surrogate marker considered likely to correspond to clinical benefit. However, the Center for Drug Evaluation and Research has concluded that two confirmatory trials failed to verify midodrine’s clinical benefit, and in August 2010 the center formally proposed to withdraw the ProAmatine NDA and midodrine ANDA approvals.

FDA Commissioner Margaret Hamburg recently accepted an agreement between CDER and Shire that allows midodrine to remain on the market while the company conducts new clinical trials (Also see "Not The Next Avastin? Plan For New ProAmatine Studies Would Negate Shire’s Withdrawal Hearing Request" - Pink Sheet, 19 Dec, 2011.).

The uncertainty surrounding midodrine’s regulatory status and continued availability was viewed as an opening for Chelsea, which is seeking approval for droxidopa based upon endpoints that measure symptoms and function (Also see "Chelsea’s Hypotension Drug Droxidopa Could Fill Void If FDA Pulls ProAmatine" - Pink Sheet, 31 Oct, 2011.).

The NDA review was expected to be challenging, in part because one of two pivotal trials failed to meet the primary efficacy endpoint. However, on Feb. 13 Chelsea signaled that the review was proving to be more problematic than anticipated.

“We were surprised by the FDA’s characterization of our safety profile” in the agency’s background briefing document, CEO Simon Pedder said.

In a press release and presentation at the Biotechnology Industry Organization’s CEO and Investor Conference on Feb. 13, Chelsea announced that FDA’s background briefing documents (which the company would have received two to three weeks in advance of the advisory committee meeting) raised several key concerns related to clinical endpoints, long-term efficacy and safety. In particular, the agency placed increased emphasis on safety data from Chelsea’s long-term extension trials and post-marketing surveillance program in Japan, where droxidopa has been approved since 1989.

“We were surprised by the FDA’s characterization of our safety profile in that briefing document,” President and CEO Simon Pedder said at the BIO conference. “It certainly has been different from prior discussions we’ve had with the agency. Subsequently, we felt strongly that the shift in the risk analysis was sufficiently material to share” publicly.

Short-Term Efficacy Seen …

FDA’s briefing documents, released to the public on Feb. 21, include a Jan. 27 memo in which clinical reviewer Melanie Blank recommends against approving droxidopa at this time “on the basis of the safety concerns compounded by absence of evidence of durability of effect.” Blank said efficacy and safety questions remain due to the brief duration of the two pivotal trials, only one of which met its primary endpoint.

Conducted under a Special Protocol Assessment, Study 301 enrolled 162 subjects in a double-blind, randomized phase following an open-label droxidopa titration period and a one-week washout. The primary efficacy assessment was based on the composite Orthostatic Hypotension Questionnaire.

OHQ comprises two subscales: the Orthostatic Hypotension Symptom Assessment scale, which includes six items that rate the presence and severity of NOH symptoms (dizziness/lightheadedness, problems with vision, weakness, fatigue, trouble concentrating and head/neck discomfort), and the Orthostatic Hypotension Daily Activities Scale, which includes four items that rate the impact of NOH symptoms on daily activities that require standing or walking.

After one week of droxidopa treatment in the double-blind period, droxidopa was associated with a statistically significant improvement in the OHQ, with a mean effect size of 0.9 points. There also was a statistically significant improvement in OHSA Item 1, which measures the cardinal NOH symptom of dizziness/lightheadedness, with a mean effect size of 1.3 points.

Study 301’s primary efficacy endpoint was changed from OHSA 1 to OHQ after an earlier trial, Study 302, failed to show a significant benefit on the former endpoint but showed a nominally statistically significant improvement in OHQ in an exploratory analysis. Study 302 was a withdrawal study in which the randomized, double-blind treatment period was 14 days.

A third study, 303, also had a randomized withdrawal design. There was an initial three-month, open-label treatment period, followed by a two-week double-blind, randomized withdrawal period and then a nine-month open-label treatment period. The study was not powered to show an effect on OHQ and did not show a significant benefit on either OHQ or OHSA 1.

… But Small Numbers Complicate Safety Evaluation

In her review, Blank noted the small percentage of patients in the droxidopa clinical program who received the drug for either an extended period of time or at the highest proposed dose. In an errata to her review, Blank said 101 of 535 patients were exposed for more than a year, and only 28 of those were exposed to the maximum dose of 600 mg three times daily.

“There was limited Phase III double-blind exposure; only 131 patients received droxidopa with a mean exposure of 11 days during the double-blind Phase III studies. This low degree of long-term exposure makes it difficult to properly evaluate the long-term safety of droxidopa,” Blank wrote. “Additionally, the limited exposure to the highest dose … makes it difficult to make a proper safety assessment.”

Given that Chelsea is pursuing an orphan indication, the small size of the safety database would not have been so problematic “except that the safety of droxidopa was not so clean,” Blank noted. “During the longer-term open label experience with droxidopa, there were several deaths, SAEs, discontinuations for AEs and events of hypertensive crisis, strokes and myocardial infarction. There were also several patients with worsening of their movement disorders.

“Of utmost concern are reports of neuroleptic malignant syndrome from Japan that aren’t clearly explained,” Blank continues. During a 10-year reporting period, there were nine case of the syndrome, an often-fatal neurological condition, reported in Japan among patients taking droxidopa.

The decision on whether to approve droxidopa “is debatable,” FDA’s clinical reviewer said.

In her memo, Blank said the decision on whether to approve droxidopa “is debatable,” and she sets out arguments both for and against approval.

Factors favoring approval included strong evidence from Study 301 that droxidopa confers at least one week of symptomatic benefit on the core NOH symptom of dizziness/lightheadedness, and on the comprehensive OHQ scale. Most of the OHQ individual component scores also showed a statistically significant benefit with droxidopa, and there was strong evidence from Study 301 that droxidopa raises standing systolic blood pressure for at least one week.

Key among the factors weighing against approval are the lack of a showing of durable effect (lasting more than four weeks) and uncertainties about the drug’s safety profile due to the pivotal studies’ design and duration.

Although Studies 302 and 303 showed slightly favorable trends for droxidopa on OHSA Item 1, they “did not demonstrate clinical/symptomatic benefit for droxidopa after two weeks and three months, respectively, of chronic use followed by a two-week randomized withdrawal period. These studies also failed to show any durable effect on systolic blood pressure,” Blank wrote.

The failure to demonstrate durability of effect is important given the various safety findings in the open-label periods and the nine cases of neuroleptic malignant syndrome in Japan, some of which do not appear to be explained on the basis of other drugs known to cause the syndrome, Blank said.

Furthermore, the effect size of 1.3 on OHSA Item 1 in Study 301 is not easily understood. “The sponsor did not provide sufficient data to allow one to equate this change in symptom score to clinical benefit.”

In its draft questions to the committee, FDA asks whether Study 301 alone could have provided sufficient evidence of treatment effect, and what impact Studies 302 and 303 had on the efficacy evaluation. The agency asks whether it is important that a symptomatic treatment for a chronic condition be shown effective for more than two weeks, whether effect size matters for a symptomatic claim and, if so, whether droxidopa’s effect size is clinically meaningful.

FDA also wants the committee to discuss the safety experience and whether Chelsea’s proposal for a post-marketing observational registry of 200-300 patients followed for four to five years would provide adequate reassurance about droxidopa’s safety. There is only one voting question: whether droxidopa should be approved.

Challenging Chelsea’s Endpoints

FDA’s draft questions do not seek the committee’s views on the validity and utility of the OHQ and its subscales. However, the review documents call into question the use of these instruments, with the exception of OHSA Item 1, which is viewed as assessing a core NOH symptom.

In a Jan. 23 memo, Study Endpoints and Labeling Development reviewer Elektra Papadopoulos said the OHSA is not a comprehensive symptom measure because it excludes imbalance and falls – symptoms that have been described in qualitative research as among the most troubling – and asks questions about extraneous symptoms.

The OHDAS also is not a comprehensive assessment of the impact of NOH on a patient’s daily activities because it does not assess those activities that require positional changes, such as going from lying to sitting to standing. Combining the OHSA and OHDAS scores into the single OHQ composite also is problematic.

“If droxidopa is approved, we do not advise including a reference to the assessment used to support treatment benefit in the Indications section of labeling,” Papadopoulos said. “Statements that name the OHQ overall score or its subscales (OHSA and OHDAS) should be avoided in all sections of labeling.”

The SEALD reviewer recommends that the concepts contained in OHSA Item 1 – dizziness, lightheadedness, feeling faint – be described in the Clinical Studies section of labeling because they are the core symptoms of NOH.