Letter to FDA Commissioner Hamburg on New Antibiotic Product (CAZ-AVI)

December 19, 2014

Margaret A. Hamburg, M.D.
Food and Drug Administration
10903 New Hampshire Ave
Silver Spring, MD 20993-0002

Re: Anti-Infective Drugs Advisory Committee (AIDAC) Meeting on December 5, 2014

Dear Commissioner Hamburg,

We are writing as members of the Patient, Consumer, and Public Health Coalition to express our strong concerns about the approval standards considered and the conflicts of interest apparent at the December 5, 2014 FDA meeting of the Anti-Infective Drugs Advisory Committee (AIDAC) regarding a new combination antibiotic product, ceftazidime-avibactam (CAZ-AVI). This antibiotic was being considered for three indications: complicated intra-abdombinal infections (cIAI), complicated urinary tract infections (cUTI), and aerobic gram-negative infections, including HABP/VABP (hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia).

All of us understand the dangers posed by antibiotic resistance and our diminishing ability to treat serious infections. However, as you know, ineffective antibiotics worsen antibiotic resistance and simultaneously expose patients to unnecessary risks while delaying administration of effective treatment. From this perspective, we were deeply troubled by the minimal evidence presented by the sponsor in support of this application, and the overall regulatory approach being taken by the FDA, including use of the section 505(b)(2) pathway for this new entity and its indication for use among a poorly defined patient group. Specifically,

— Limited clinical experience with CAZ-AVI did not meet the FDA standard of substantial evidence for any indication

    • There was Phase III level data only for one indication (cIAI), and those data were preliminary data from one trial, which was not reviewed by the FDA prior to the Advisory Committee meeting. It is not clear why the Advisory Committee meeting was not rescheduled to be held after the FDA had time to review and it is shocking that FDA scientists made it clear in the briefing material that FDA review of the data was not essential prior to approval.
    • The evidence for cUTI and remaining evidence for cIAI included preliminary interim results from one ongoing pooled open-label trial, two small phase II trials and a meta-analysis, all of which had substantial inadequacies, described in detail in this letter. This is completely inadequate and should not have been considered sufficient to be considered for recommending approval by an Advisory Committee.
    • There was absolutely no human data to support the last indication, for aerobic gram-negative infections, including HABP/VABP (hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia). We strongly believe that including this untested indication for review by an Advisory Committee sets a terrible precedent for companies seeking approval for their drugs.

 Submitted studies all had significant limitations identified by the FDA

    • We agree with the FDA scientists regarding the interim results of the open-label trial for cIAI and cUTI, who stated “there are concerns about exchangeability of subjects due to potential differences in doses and infusion time, baseline patient and disease characteristics, prognostic factors and the supportive care they had received.”
    • The FDA briefing material described the phase II trials for cIAI and cUTI as having “no pre-specification of any formal hypotheses for inferential testing, and statistical analysis is limited to descriptive data summaries.” The FDA briefing material described the sponsor’s meta-analysis as having “considerable uncertainties in these studies…they are not similar with respect to design, dose, and duration of treatment, baseline disease characteristics, timing of assessment and other factors.” We agree.

Taken together, these studies do not constitute the substantial evidence of safety and efficacy required by the FDA. Scheduling an Advisory Committee meeting to even consider three indications with so little scientific evidence to support them sends a message that the FDA has drastically lowered its bar for the evidence needed for approval. It is disturbing that most Advisory Committee members did not seem to understand the low quality of these data and how risky it would be to use these unproven drugs, and recommended approval for two out of the three indications being sought (for both cIAI and cUTI). This raises questions regarding the limited scientific expertise of these Committee members, most of whom are primarily clinicians. It may be that the data on the first 2 indications were perceived as adequate only because they were better than the data for the 3rd indication, which received a unanimous rejection from Committee members.

— Safety and efficacy concerns were identified in the preliminary analysis of the phase III study; the study has not yet been reviewed by the FDA

    • The preliminary results of the phase III trial for cIAI had 8 deaths (26%) in the CAZ-AVI + MTZ group versus only 3 deaths (9%) in the comparator group, among patients with moderately impaired renal function. These patients, who are a significant clinical demographic, also had a lower clinical cure rate (45% in the CAZ-AVI group versus 74% in the comparator group). The sponsor stated that the mortality imbalance was unlikely to be due to the study drug and that the observed lack of efficacy is likely due to drug underexposure. That is a biased interpretation of scientific data, given that the briefing material from both the FDA and the sponsor indicated that the small size of this subgroup made these findings difficult to interpret. Follow-up studies are needed to address these concerns before patients with compromised renal function are exposed to the drug. Otherwise, these patients may be put needlessly at risk when more effective treatment options already exist. It was very disappointing that the committee voted 11 to 1 in favor of recommending approval for this indication despite the paucity of evidence. While this was the only indication with Phase III data, the results were very worrisome and the FDA should reject the Advisory Committee’s ill-considered recommendation.

 Questionable use of the section 505(b)(2) pathway

    • The FDA background document stated that “The Agency’s regulatory approach to this NDA is to rely on the Agency’s prior finding of safety and effectiveness of ceftazidime. The contribution of avibactam will be demonstrated by in vitro studies, animal models of infection and with limited clinical data. Safety of avibactam will be determined from Phase 1 studies of avibactam alone and phase 1 and 2 studies of CAZ-AVI.”

This combination still represents a novel drug entity with unique properties; therefore, we find application of the section 505(b)(2) pathway here inappropriate. The FDA background document asserts that “the avibactam component is a new chemical entity that is not currently marketed in any country, either alone or in combination.” In the 1970 Upjohn versus Finch court case, it was determined that for “…combination drugs purported to have advantages exceeding those of the components, there must be adequate, well-controlled data documenting the claimed advantages.”[1] As the mechanism-of-action of avibactam is to restore activity of ceftazidime among certain ceftazidime-resistant pathogens, CAZ-AVI meets the definition described in Upjohn versus Finch, and renders invocation of the 505(b)(2) pathway here dubious at best.

— Poorly defined indication for use among those with “limited or no alternative treatment options” could easily result in dangerous overuse

The last indication being sought by the sponsor was described in the FDA briefing materials as for “aerobic gram-negative infections with limited or no alternative treatment options including HABP/VABP.” However, in the sponsor’s presentation at the meeting, their wording of this indication was exclusively “no alternative treatment options.” This discrepancy was noted by the advisory committee and prompted the use of an additional voting question to separately address each of these possible wordings. This action by the Advisory Committee highlights the importance of clearly defining the intended patient population. The term “limited options” is vague and is likely to be used to justify exposure of patients to treatments that have increased risks when safer and more effective treatments are available.

 Potential conflict of interest not disclosed during the open public hearing

Open public hearing speakers are encouraged to share any conflicts of interest with the committee at the beginning of their testimony. At the December 5 meeting and other meetings of this committee, we have observed speakers who state that they have no conflicts of interest, but they don’t mention that they represent organizations that have substantial financial ties to companies with conflicts of interest. In addition, IDSA is treated by FDA as an independent voice, when in fact their membership has significant ties to the companies that make antibiotics. In the interest of a transparent meeting process, the FDA should encourage speakers to disclose not only individual conflicts of interest, but also potential conflicts of interest for organizations that they represent.


  1. All studies submitted by the sponsor had serious limitations and do not meet the evidential standard of the FDA.
  2. The regulatory approach taken by the FDA is based on improper use of the 505(b)(2) pathway and poorly defined patient populations. Because they have not clearly defined the target population, this is likely to result in needless risk for patients.
  3. There were NO studies in humans to support the indication for aerobic gram-negative infections, including HABP/VABP (hospital-acquired bacterial pneumonia and ventilator-associated bacterial pneumonia). This indication was rejected by the Advisory Committee but should not have even been included for consideration.
  4. Presenters at open public hearings should be encouraged to disclose both individual and organization-level conflicts of interest.

We strongly urge the FDA to require the sponsor to complete all ongoing phase III studies, as well as its own review, before considering CAZ-AVI for approval for any indication. Indications which are intended for patients with few existing treatment options should be more clearly defined in order to prevent use of the drug by patients who are more likely to be harmed than helped by a drug approved only for a narrow targeted population.

American Medical Student Association
Annie Appleseed Project
Breast Cancer Action
Center for Medical Consumers
Connecticut Center for Patient Safety
Jacobs Institute of Women’s Health
National Center for Health Research
National Women’s Health Network
Our Bodies Ourselves
The TMJ Association

Contact Information: Anna Mazzucco, PhD at (202) 223-4000 or am@center4research.org

[1] Upjohn v Finch, 422 F 2d 944. 1970.