6.1. Commercial Development of an Anthrax Antidote
With regard to developing medical countermeasures for treatment of anthrax disease, Human Genome Sciences (HGS) announced the following in press releases or in the briefing package for the October 2009 meeting of the Vaccines and Related Biological Products Advisory Committee [23
- October 2002: pre-IND meeting between HGS and FDA;
- February 2003: use of protein and antibody drug development capabilities to develop therapeutic candidates to address microbial targets including anthrax;
- March 2003: discovery of a human monoclonal antibody drug that is effective in protecting against anthrax in multiple experimental models in animals;
- May 2003: submitted IND to FDA;
- June 2003: clearance from the Food and Drug Administration (FDA) to begin human trials of raxibacumab;
- July 2003: initiation of clinical development of raxibacumab for the prevention and treatment of anthrax infections;
- August 2003: designation from the FDA for raxibacumab as a Fast Track Product;
- November: orphan drug designation for raxibacumab granted by FDA;
- March 2004: finding in the Phase 1 trial that raxibacumab is safe and well tolerated in healthy volunteers and achieved the blood levels predicted by relevant animal models as necessary to afford significant protection from the lethal effects of the anthrax toxin;
- 2004 to 2007: a series of meetings with FDA regarding the additional animal efficacy studies needed to support licensure and/or use in the Strategic National Stockpile. These interactions established;
- the requirement to demonstrate efficacy in two species (rabbits and monkeys);
- that the animals had to have evidence of systemic anthrax disease at the time of raxibacumab administration for an indication in therapeutic treatment;
- that serum protective antigen (PA) could be used as a trigger for therapeutic treatment;
- that the antibiotic exposure in animals in the raxibacumab/antibiotic combination studies should approximate the exposure achieved by the recommended dose in humans.
Agreement on the division of studies between those needed to support submission of an IND by the CDC to use raxibacumab in the Strategic National Stockpile and those additional studies needed for licensure was also achieved. During 2007, HGS submitted the protocols and analysis plans for the rabbit and monkey efficacy studies. The protocols for the rabbit and monkey raxibacumab/antibiotic combination studies were completed in the summer of 2008.
- July 2005: publication of Phase 1 study results in Clinical Infectious Diseases;
- October 2005: award for a two-phase contract to supply raxibacumab to the United States Government. Under the first phase of the contract, HGS would supply ten grams of raxibacumab to the Department of Health and Human Services for comparative in vitro and in vivo testing. Under the second phase of the contract, the Government had the option to order up to 100,000 doses of raxibacumab for the Strategic National Stockpile;
- June 2006: government exercised option to purchase 20,000 treatment courses of raxibacumab for the Strategic National Stockpile;
- October 2008: Pre-Biologics License Application meeting with FDA;
- February 2009: commencement of delivery of 20,000 doses raxibacumab to the Strategic National Stockpile;
- May 2009: submission of a Biologics License Application (BLA) for raxibacumab;
- July 2009: publication in The New England Journal of Medicine of the results of two pivotal animal efficacy studies, which showed the life-saving potential of raxibacumab in the event of life-threatening inhalation anthrax disease;
- July 22, 2009: Government exercised option to purchase additional 45,000 doses of raxibacumab for the Strategic National Stockpile;
- October 2009: consideration of the BLA for raxibacumab at FDA's Anti-Infective Drugs Advisory Committee;
- November 2009: FDA issued Complete Response Letter requesting additional information relating to the BLA for raxibacumab;
- March 2010: $180.2 million in sales of raxibacumab to the Strategic National Stockpile in 2009 (the first product sales for the company);
- January 2011: HGS working closely with FDA to obtain approval of raxibacumab for the treatment of inhalation;
- February 2011: $47.2 million in sales of raxibacumab to the Strategic National Stockpile in 2010.
6.2. FDA Advisory Committee
In October 2009, FDA's Anti-Infective Drugs Advisory Committee convened to consider the Biologics License Application from Human Genome Sciences for raxibacumab injection, a monoclonal antibody product for treatment of inhalation anthrax [25
]. At the outset of the meeting, FDA indicated that results of inspections of bioanalytical assays for raxibacumab and ciprofloxacin raised questions about the reliability of the clinical pharmacology data. Therefore, FDA would discuss neither pharmacokinetic or pharmacodynamic results nor the selection of a human dose.
6.3. Human Genome Sciences' Perspective
Dr. Sally Bolmer of Human Genome Sciences presented results of anthrax challenge studies in New Zealand white rabbits and cynomolgus monkeys, both well-characterized models of inhalational anthrax. Raxibacumab doses of 20 and 40
mg/kg achieved concentrations that were at least as high as the highest PA levels in anthrax-infected animals. The pivotal animal efficacy studies had parallel groups, with animals randomized to raxibacumab or placebo. To mimic the human clinical situation, other studies were conducted in which animals in some of the treatment groups received antibiotics at the onset of symptoms after anthrax challenge. Dr. Bolmer noted, however, that the preamble to the Animal Rule describes the need for a wide range of therapeutic options for the treatment of bioterror pathogens and specifically cites anthrax. She noted further that the preamble states that there is no requirement for new therapies to provide meaningful therapeutic benefits to subjects over existing therapies, nor does it require the toxic agent to be without a proven treatment. Finally she reported that concomitant administration of raxibacumab with antibiotics in animals did not alter the antibiotic efficacy or pharmacokinetics. In other words, raxibacumab was effective in the animal models, and antibiotics were effective in the animal models, and there was no decrease in effectiveness when raxibacumab was administered with antibiotics. Dr. Bolmer concluded her remarks by introducing Dr. Daniel Lucey, who was chief of the Infectious Disease Service at the Washington Hospital Center in Washington, DC, during the anthrax attacks in 2001.
Dr. Lucey noted that in controlled experimental settings, antibiotics can achieve up to 100% survival but that in real world clinical use antibiotics are not as successful. He noted that in the United States in the 20th century, mortality from inhalational anthrax was 90% with antibiotic susceptible strains of anthrax [26
]. In the anthrax attacks in 2001, survival was 55% in the 11 patients with inhalational anthrax, and Dr. Lucey considered that more rapid blood culture results, use of two or more antibiotics (ideally at least one which crosses the blood-brain barrier in order to prevent or treat anthrax meningitis), and pleural fluid drainage have improved survival. Nevertheless, the mortality rate of 45% was highly unacceptable, and he considered that an antitoxin is needed as an additional treatment modality, because the toxins still exert deleterious effects after control of bacterial replication with antibiotics.
Dr. Lucey went on to say that prior to the development of symptoms, antibiotics were very effective for postexposure prophylaxis in 2001. Even in the first one to five days of prodromal flu-like symptoms, prompt treatment with antibiotics may have great benefit. In an inhalational anthrax infection, the prodromal phase is followed by an intermediate progressive phase characterized by bacteremia and/or pleural effusions and/or mediastinal adenopathy. Lastly, progression to the late fulminant stage can occur rapidly and lead to death within 6 to 24 hours. At this late fulminant stage, there has been no demonstrated benefit from antibiotics alone, and all five patients in 2001 who died had reached this stage. Finally, Dr. Lucey noted that the treatment paradigm of using an antitoxin in combination with antimicrobials is well established for tetanus.
Dr. Thi-Sau Mignone of Human Genome Sciences described the pivotal animal studies with raxibacumab in detail. The challenge dose of anthrax was 200 times the LD50
(the dose that causes lethality in 50% of the animals). At this dose, the majority of rabbits died between days 3 and 5 whereas the majority of monkeys died between days 4 and 6. Positive bacteremia and anthrax PA in blood occurred at 24 hours in the rabbits and at 36 hours in the monkeys. A significant rise in temperature was also seen in rabbits at 24 hours, but monkeys have a strong diurnal rhythm, and temperature rise was not observed at 36 hours. When raxibacumab was administered at 40
mg/kg at the onset of these symptoms, 44% of the rabbits and 64% of the monkeys survived.
The studies in which antibiotics and raxibacumab were coadministered upon onset of symptoms, as would be expected in clinical practice in humans, were conducted to test whether raxibacumab would interfere with levofloxacin or ciprofloxacin. The studies were not designed to detect superiority or noninferiority of the combined treatment to treatment with the antibiotic alone but rather to detect superiority of the combined treatment to placebo. In both rabbits and monkeys, survival was statistically significant higher with combined raxibacumab and levofloxacin (rabbits) or ciprofloxacin (monkeys) than with placebo, but survival was not different between combined treatment and treatment with the antibiotic alone.
In contrast to the animal models, in humans in the anthrax attacks of 2001, symptoms developed 4 to 6 days after exposure, and death occurred 5 to 8 days after the appearance of symptoms, for a total time to death in the range of 9 to 14 days.
Dr. Dan Hanfling, special advisor for emergency preparedness and response for the Inova Health System and clinical professor of emergency medicine at George Washington University, reviewed the risk-benefit profile of raxibacumab. He had been integrally involved in the emergency management response during the 2001 attacks in the national Capitol region, during which time two postal workers were successfully diagnosed and treated. He noted that over 33,000 people were treated with prophylactic antibiotics after the 2001 attack and reported that the Commission on the Prevention of Weapons of Mass Destruction had recently warned that anthrax spores released by a crop duster could kill more Americans than died during all of World War II. Although Human Genome Sciences had not tested antibiotic-resistant strains of anthrax, he pointed out that PA is highly conserved in all known strains of anthrax and that raxibacumab would thus as a single agent be expected to be effective against antibiotic-resistant strains. He concluded that it is reasonable to predict that the benefit to risk profile in humans is strongly positive for raxibacumab as a treatment for anthrax infection.
6.4. FDA's Perspective
Dr. Yuliya Yasinskaya, a medical officer in the Division of Special Pathogens and Transplant Products at the Center for Drug Evaluation and Research, presented an FDA perspective on the raxibacumab animal studies. She pointed out that the animal studies were subject to several limitations in predicting response in humans, among which is the time of intervention, which could be different between animals in a controlled research environment and humans in a clinical setting. Antibiotics were very effective in the animal studies, with close to 100% survival.
FDA was also quite concerned about the unexpected finding of a greater incidence and severity of central nervous system (CNS) lesions in raxibacumab-treated animals that died compared to placebo animals that died. Dr. Yasinskaya noted that although only a single rabbit and a single monkey died in the combined treatment groups, there were high grade CNS lesions in 100% of those animals, whereas in the placebo group the incidence of high grade lesions was low. Dr. Yasinskaya concluded that the question remains whether in humans the addition of raxibacumab added to antibiotics will provide either additional benefit or additional risk. Due to limitations of the animal studies, it remained uncertain how long antibiotic treatment can be delayed and thus how late raxibacumab might be able to provide improved survival in these models.
6.5. Advisory Committee's Perspectives
After much preliminary discussion, the FDA questions were read and the advisory committee voted as follows.
- Does the evidence from the animal models evaluating raxibacumab at 40 mg/kg IV predict response for treatment of humans with inhalational anthrax disease? And if not, what additional studies should be conducted?
One committee member interpreted the question to refer to monotherapy. Another committee member noted that the question could be taken to mean response at any time postexposure or at any stage of clinical illness.
The committee voted 16 in favor, 7 opposed, with 1 abstention. Several of the committee members, whether they voted in favor or opposed, expressed a desire to see data from animal studies in which antibiotic treatment was not 100% successful, either because doses would be more in keeping with human doses or because timing of administration would be delayed, so that additional benefit of raxibacumab could be observed if additional benefit indeed exists. It was noted that in a disaster situation, humans would probably show up a week or more after their exposure, after they probably have been PA positive for a while, a very different situation than the animal studies.
One committee member remarked that the wording of the question was almost Talmudic, and the people who voted in favor or opposed often had all the same reasons and all the same concerns but wound upcoming down on one side or the other of the fence. One committee member who voted in favor commented she thought elements 1 through 3 of the Animal Rule were met and that we should move forward with this kind of drug. Several other committee members who voted in favor stated that they narrowly interpreted the question as referring to monotherapy with raxibacumab. One committee member explained his opposition as due to the wording of the question with respect to anthrax disease, pointing out that appearance of PA in the blood of animals is not anthrax disease.
(2) Does the evidence provided in these studies support the conclusion that raxibacumab will not diminish the anticipated efficacy of antimicrobials in inhalational anthrax? And if not, what additional studies should be conducted?
One committee member immediately noted that FDA had instructed the committee to ignore the pharmacokinetic data, because of questions about their reliability. Therefore, he asked how is the committee to answer this question if ignoring the pharmacokinetic data? FDA suggested that he should vote as he believed he could vote and then provide his perspective as to why he took that vote and how he might have voted depending on what kind of clinical data were available.
The committee voted 10 in favor, 11 opposed, with 3 abstentions. The first committee member called upon to explain his vote noted that there was a major control arm missing, raxibacumab alone, so that it was very difficult for him to make that comparison. Another who voted in opposition said that the antibiotics had been given in doses so far above their therapeutic thresholds that if there was a detrimental effect of raxibacumab, it might not have been seen in these studies. One committee member voted in favor because he knew of no precedent for a protein therapeutic like an antibody to alter the behavior of a small molecule antimicrobial. Another who voted in favor said that there was no evidence to the contrary.
The statistician committee member calculated that it would take about 400 animals to detect a statistically significant difference between 95% survival in one group versus 85% in the other. He considered that a study with that number of animals sounded like a nonstarter. One member said he abstained because he could not decide which way to vote and that the committee vote had validated his feelings. Eventually, one member said that he voted in opposition because of the narrow construction of what was presented and that he voted in favor (laughter noted in the transcript at this point) because of the narrow construction that the percentages were the same. Finally one member said that the last time he looked in a biochemistry textbook, antibiotics can be protein bound and that there was a serious matter here and that the animals should be treated with antibiotics at doses that we would use to treat a human being.
(3) Should evidence be requested that raxibacumab makes a contribution to the efficacy over the antimicrobial alone in rabbit and monkey models? If yes, what types of additional studies should be requested and then conducted?
The committee voted 17 in favor, 6 opposed, with 1 abstention. A common sentiment was that in additional studies, the animals should be given the study material at a time that is truly therapeutic and not prophylactic. One member who voted in favor noted that the question did not say who would conduct the studies and he was not sure that the “poor sponsor” who had the misfortune of developing this compound should have to do all those studies. Another member who voted in favor thought that it might be a difficult bar for the sponsor to meet. A member who voted in opposition pointed out that raxibacumab may have benefit as monotherapy if it is dosed correctly, although it would be good to know if it could help rescue patients that had delayed antibiotic therapy. Another member who voted in opposition said it may be an unreasonable burden to say raxibacumab is going to benefit every patient who receives it. He thought there was every reason to think raxibacumab is safe and efficacious and would benefit a significant fraction of those who receive it.
(4) Do you have any recommendations how this (the CNS findings) might be further evaluated?
Despite a number of committee members advocating further studies in a meningitis model, one member pointed out that their understanding was biased by the fact that the only data were from nonsurviving animals. He noted that the only way to determine whether there was causality would be to look at survivors as well to see if they have any CNS toxicity. A couple of committee members were concerned about antibody complexes as potentially a source of toxicity.
(5) Are there additional comments or further recommendations for safety evaluation in humans? If yes, what are these recommendations?
One committee member recommended that some work needs to be done to make sure that raxibacumab would be safe for use in children, and another recommended that the same level of assurance was needed for elderly patients.