animal rule for drug approval creates a jungle of confusion

2
NEWS 118 VOLUME 19 | NUMBER 2 | FEBRUARY 2013 NATURE MEDICINE Animal rule for drug approval creates a jungle of confusion In 2002, in the aftermath of the 9/11 attacks, the US introduced the ‘animal rule’ as a way for the country’s regulators to approve medical countermeasures to chemical, biological and radiological weapons in instances where conducting human efficacy studies would be unethical or unfeasible. In the rule’s first ten years, the US Food and Drug Administration (FDA) has used the provision to approve vitamin B12 injections for cyanide poisoning, an enzyme-blocking pill for nerve gas exposure and a broad- spectrum antibiotic to treat plague. Notably, in all of these cases the drugs had already received FDA approval for other afflictions or had previously received a green light in other countries for use against deadly chemicals. Now, in a regulatory first, on 14 December the FDA cleared a completely novel product through the animal rule without any such prior approvals: an antibody drug called raxibacumab for the treatment or prevention of respiratory anthrax, caused by inhalation of the bacterium Bacillus anthracis. In fact, although the safety of raxibacumab was demonstrated in more than 300 healthy human volunteers, the only validated efficacy studies available to the FDA were in monkeys and rabbits. “It’s really the first drug that has been approved using the animal rule as the sole mechanism,” says James Swearengen, director of comparative medicine at the National Biodefense Analysis and Countermeasures Center at Fort Detrick, a government research lab in Frederick, Maryland, created by the US Department of Homeland Security. “It’s a milestone.” For Swearengen, the decade-plus it took from enacting the animal rule to approving raxibacumab is simply a reflection of the pace of regulatory science: “Any time you introduce anything new like that at a government level, there’s a learning curve,” he says. However, others worry that the dearth of medical countermeasures approved in the post-9/11 era signals a deeper problem— both with the FDA’s handling of applications under the animal rule and with the rule itself. “While it’s great that we have this drug that’s approved, the stockpile is not as full as I would like it to be,” says Gigi Gronwall, an immunologist at the Center for Biosecurity of the University of Pittsburgh Medical Center in Baltimore. Both Swearengen and Gronwall sat on a committee first convened in 2009 by the US National Academy of Sciences (NAS) that was tasked with judging how reliable animal models are as proxies when it comes to testing the efficacy of bioterrorism countermeasures. Under the existing animal rule, the FDA will accept animal efficacy data when a protective effect is demonstrated in at least two suitable species or in a single animal species if it “represents a sufficiently well-characterized animal model for predicting the response in humans.” Yet, according to Thomas Hartung, an NAS panel member and director of the Center for Alternatives to Animal Testing at the Johns Hopkins Bloomberg School of Public Health in Baltimore, such an ideal model simply doesn’t exist. “If there was an animal model good enough to substitute for people, we would not have a 92% failure rate in clinical trials,” he says. Take sepsis, for example. This life- threatening condition has a similar clinical profile of inflammation gone haywire as that seen in people who are infected with anthrax, Ebola virus, botulism or other bioterrorism agents. Yet, despite the fact that some compounds have repeatedly reversed the symptoms of sepsis in animal tests, not a single drug has proven effective in human clinical trials, even though more than 30,000 people have been included in randomized control studies involving candidate antisepsis agents over the past 25 years (see Nat. Med. 18, 997–1002, 2012). Why, then, scientists wonder, should one or two animal models be seen as sufficient for an anthrax drug, for example, when there has been such difficulty with sepsis trials? “The bottom line,” Hartung says, “is that we can’t recommend any animal model, and it’s not about refining those we have.” A chip off your shoulder To relieve the pressure to find optimal animal models, the NAS committee, in its report released in December 2011, recommended adopting a concept of “compartmentalization.” Under this approach, efficacy data can be gleaned from components of an animal—organs, cells or biological networks, such as the immune system—rather than relying on information obtained solely from whole organisms. Hartung also advocates developing more predictive in vitro systems that combine stem cells with microfluidics technologies to create three-dimensional human organ equivalents. And he’s not alone. The US Department of Defense and the US National Institutes of Health recently pledged a total of close to $200 million over five years to advance what they’re calling the ‘human on a chip’ approach. Although the FDA is supporting the development of such chip-based technologies, the agency has not given up on animal models entirely. In 2011, it launched an Animal Model Qualification Program to help researchers develop animal systems and validate how well they mirror specific human ailments. Additional clarity on animal rule requirements is likely to come later this year Under the hood: GLP-compliant studies prove a challenge in high-containment laboratories. Texas A&M npg © 2013 Nature America, Inc. All rights reserved.

Upload: elie

Post on 14-Dec-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Animal rule for drug approval creates a jungle of confusion

n e w s

118 volume 19 | number 2 | february 2013 nature medicine

Animal rule for drug approval creates a jungle of confusionIn 2002, in the aftermath of the 9/11 attacks, the US introduced the ‘animal rule’ as a way for the country’s regulators to approve medical countermeasures to chemical, biological and radiological weapons in instances where conducting human efficacy studies would be unethical or unfeasible. In the rule’s first ten years, the US Food and Drug Administration (FDA) has used the provision to approve vitamin B12 injections for cyanide poisoning, an enzyme-blocking pill for nerve gas exposure and a broad-spectrum antibiotic to treat plague. Notably, in all of these cases the drugs had already received FDA approval for other afflictions or had previously received a green light in other countries for use against deadly chemicals.

Now, in a regulatory first, on 14 December the FDA cleared a completely novel product through the animal rule without any such prior approvals: an antibody drug called raxibacumab for the treatment or prevention of respiratory anthrax, caused by inhalation of the bacterium Bacillus anthracis. In fact, although the safety of raxibacumab was demonstrated in more than 300 healthy human volunteers, the only validated efficacy studies available to the FDA were in monkeys and rabbits.

“It’s really the first drug that has been approved using the animal rule as the sole mechanism,” says James Swearengen, director of comparative medicine at the National Biodefense Analysis and Countermeasures Center at Fort Detrick, a government

research lab in Frederick, Maryland, created by the US Department of Homeland Security. “It’s a milestone.”

For Swearengen, the decade-plus it took from enacting the animal rule to approving raxibacumab is simply a reflection of the pace of regulatory science: “Any time you introduce anything new like that at a government level, there’s a learning curve,” he says. However, others worry that the dearth of medical countermeasures approved in the post-9/11 era signals a deeper problem—both with the FDA’s handling of applications under the animal rule and with the rule itself. “While it’s great that we have this drug that’s approved, the stockpile is not as full as I would like it to be,” says Gigi Gronwall, an immunologist at the Center for Biosecurity of the University of Pittsburgh Medical Center in Baltimore.

Both Swearengen and Gronwall sat on a committee first convened in 2009 by the US National Academy of Sciences (NAS) that was tasked with judging how reliable animal models are as proxies when it comes to testing the efficacy of bioterrorism countermeasures. Under the existing animal rule, the FDA will accept animal efficacy data when a protective effect is demonstrated in at least two suitable species or in a single animal species if it “represents a sufficiently well-characterized animal model for predicting the response in humans.” Yet, according to Thomas Hartung, an NAS panel member and director of the Center for Alternatives to Animal Testing at the Johns Hopkins Bloomberg School of

Public Health in Baltimore, such an ideal model simply doesn’t exist. “If there was an animal model good enough to substitute for people, we would not have a 92% failure rate in clinical trials,” he says.

Take sepsis, for example. This life-threatening condition has a similar clinical profile of inflammation gone haywire as that seen in people who are infected with anthrax, Ebola virus, botulism or other bioterrorism agents. Yet, despite the fact that some compounds have repeatedly reversed the symptoms of sepsis in animal tests, not a single drug has proven effective in human clinical trials, even though more than 30,000 people have been included in randomized control studies involving candidate antisepsis agents over the past 25 years (see Nat. Med. 18, 997–1002, 2012).

Why, then, scientists wonder, should one or two animal models be seen as sufficient for an anthrax drug, for example, when there has been such difficulty with sepsis trials? “The bottom line,” Hartung says, “is that we can’t recommend any animal model, and it’s not about refining those we have.”

A chip off your shoulderTo relieve the pressure to find optimal animal models, the NAS committee, in its report released in December 2011, recommended adopting a concept of “compartmentalization.” Under this approach, efficacy data can be gleaned from components of an animal—organs, cells or biological networks, such as the immune system—rather than relying on information obtained solely from whole organisms. Hartung also advocates developing more predictive in vitro systems that combine stem cells with microfluidics technologies to create three-dimensional human organ equivalents. And he’s not alone. The US Department of Defense and the US National Institutes of Health recently pledged a total of close to $200 million over five years to advance what they’re calling the ‘human on a chip’ approach.

Although the FDA is supporting the development of such chip-based technologies, the agency has not given up on animal models entirely. In 2011, it launched an Animal Model Qualification Program to help researchers develop animal systems and validate how well they mirror specific human ailments.

Additional clarity on animal rule requirements is likely to come later this year Under the hood: GLP-compliant studies prove a challenge in high-containment laboratories.

Texa

s A

&M

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.

Page 2: Animal rule for drug approval creates a jungle of confusion

n e w s

nature medicine volume 19 | number 2 | february 2013 119

when the FDA releases an updated version of its draft guidance on the animal rule that was first published in January 2009. That document should help expedite the development of medical countermeasures by clarifying the path to approval. “There’s tremendous confusion,” says Travis Warren, a research scientist at the US Army Medical Research Institute of Infectious Diseases at Fort Detrick who studies antiviral drugs for Ebola, Marburg and other pathogens and who can’t wait for those guidelines to come out. “We’re very much looking for additional guidance.”

One of the points of confusion centers on what experiments need to be conducted under ‘good laboratory practice’ (GLP) conditions. The GLP data-quality system was designed to ensure the consistency and reproducibility of preclinical studies included in FDA applications by requiring extensive documentation of everything involved in an experimental test. The level of detail required is so substantial, in fact, that the cost of generating GLP-compliant documentation often exceeds the cost of the research itself. The rules were written with toxicology and pharmacokinetic animal studies in mind. But under the animal rule as written, all studies intended to support drug approval, including the pivotal efficacy studies, which combine drug administration and pathogen exposure in an animal to test whether a drug would be expected to work in afflicted humans—something that would never occur in a routine drug safety test—must be conducted under GLP regulations.

That’s a tall order when working in a high-containment environment. And according to Warren, who has conducted GLP studies in a biosafety level 4 (BSL4) facility, it’s led to a lot of guesswork on the part of drug developers working in the countermeasure arena as to what the FDA will want to see in a final application. “The GLPs were not written with efficacy studies in mind,” he says. “So, from an efficacy standpoint, researchers just have to say, ‘Okay, well, what we think will be needed in this situation is something like this.’”

Over the past six months or so, Rosemary Roberts, director of the Office of Counter-Terrorism and Emergency Coordination at the FDA’s Center for Drug Evaluation and Research in Silver Spring, Maryland, has begun speaking publically to clarify the agency’s position on GLP requirements. “We do not require GLP,” she told Nature Medicine. “That said, the sponsor needs to provide a plan as to how they’re going to assure the quality and integrity of the data.

Clearly, the good laboratory practices [for preclinical animal studies] and also the good clinical practices that are used for human studies are good examples of systems and programs that do that.”

In April, the Galveston National Laboratory in Texas, with funding from the FDA, plans to host a five-day academic training course outlining best practices for how to best follow Roberts’s advice in a BSL4 setting. “This is really something to improve understanding across the board,” says Trevor Brasel, study director for regulated studies at the University of Texas Medical Branch in Galveston, which hosts the laboratory.

Raxibacumab on the rackNobody knows the need for greater guidance more than the Maryland-based company behind raxibacumab, Human Genome Sciences (HGS), which was bought by the UK’s GlaxoSmithKline last year for $3.6 billion. After having demonstrated that raxibacumab provides a survival advantage in rat, rabbit and monkey models of anthrax exposure and that the drug is safe in people, in 2007 HGS, in consultation with the FDA, commenced the pivotal GLP efficacy studies designed to support the drug’s licensure under the animal rule. Those studies showed that a particular dose of raxibacumab, a monoclonal antibody directed against a component of the anthrax toxin, protected 8 of 18 rabbits and 9 of 14 monkeys from the sure death experienced by all placebo-treated animals after inhalational anthrax exposure (N. Engl. J. Med. 361, 135–144, 2009).

But when the FDA’s Anti-Infective Drugs Advisory Committee considered raxibacumab for approval in October 2009, the panel insisted on additional studies to demonstrate whether or not the drug offered additional benefit beyond that provided by antibiotics—something the agency hadn’t asked for when it signed off on the study protocols two years earlier. HGS, with financial backing from the Biomedical Advanced Research and Development Authority (BARDA), recently finished those studies, showing that 32 of 39 rabbits treated with levofloxacin and raxibacumab survived exposure to anthrax compared with 24 of 37 rabbits that received the antibiotic treatment alone. Ten years after the first meetings took place between HGS and the FDA to discuss raxibacumab, the drug won approval.

That handling of HGS’s application has left Paul Aebersold, a former FDA staffer who now runs his own consulting company in Washington, DC, with a negative impression of the animal rule. “They basically gave the

company an incredibly bad deal by not communicating what they wanted—or maybe they didn’t know what they wanted,” says Aebersold, who has studied the raxibacumab case (Adv. Prev. Med. 2012, 507571, 2012). “HGS did everything they could to work with the FDA, and then the FDA stabbed them in the back. It was really, really shoddy.”

According to Maureen Hardwick, executive secretary of the Alliance, a Washington, DC–based coalition of drug companies working in the countermeasure field (including HGS), the lines of communication between the FDA and drug companies developing countermeasures are now much more open. “From the commissioner on down, there’s a strong message in the FDA that they want to find a way to work through these products,” she says. “There seems to be a real shift in the momentum. People feel that they are getting better information earlier from the FDA.”

For all the uncertainties surrounding the rule, many companies are still hoping their candidate drugs will meet the animal rule’s unique stipulations. For example, in September 2012, Cangene, based in Winnipeg, Manitoba, applied for US approval of an  investigational botulinum antitoxin containing a horse-derived antibody that neutralizes all seven known botulinum toxin types. The company did so by demonstrating efficacy in two animal models, with human safety information from two phase 1 clinical trials and some real-world data from people who received the drug on an emergency basis from the US Centers from Disease Control and Prevention after exposure to the botulism pathogen, mostly through spoiled food.

“We’ve come a long way to developing the infrastructure and natural history profiles for each of the present threats,” says BARDA director Robin Robinson. “Right now, we’re on the right path to getting many of those products approved under the [animal] rule as it’s been written and executed.” But, as many experts point out, one might imagine many more drugs would be in the pipeline if the rule was even clearer.

Elie Dolgin

Anthrax antidote: Antibody drug wins approval.

Hum

an G

enom

e Sc

ienc

es

npg

© 2

013

Nat

ure

Am

eric

a, In

c. A

ll rig

hts

rese

rved

.