 |
DIVISION OF ANTI-INFECTIVE DRUG PRODUCTS
POINTS TO CONSIDER
CLINICAL DEVELOPMENT AND LABELING OF
ANTI-INFECTIVE DRUG PRODUCTS
1992 ADDENDUM
DOSE-RESPONSE TESTING
One of the primary goals of drug development is to develop data that permit physicians to select the appropriate starting dose of the drug and subsequently to adjust dosage to the needs of a particular patient or group of patients. Ideally such knowledge should allow the physician to increase the likelihood of achieving the intended therapeutic effect while reducing the risk of causing adverse events. One clinical development approach utilized with other drug classes to achieve this goal has been prospective, randomized, multi-dose-level clinical parallel dose-response testing. However, it has been the experience of the Division that few applicants have approached the clinical development of antimicrobial drug products from this particular perspective.
The Division would be particularly keen on exploring with applicants the feasibility of such an approach with antimicrobial drug products. We are aware such a development approach requires a minimum of three, and, ideally, more dosing regimens to begin to characterize the shape and location of the dose response curve from which clinical dosing recommendations can be deduced. Likewise we are cognizant of the fact that, for this approach to work, some group or groups in this randomized trial scheme would most likely receive an inherently sub-therapeutic dosing regimen or a regimen on which most of the individuals would likely be therapeutic failures. The ethical problems presented in such a trial design involving individuals with infections that could otherwise be treated effectively do not go unnoticed. Likewise the ethical problems of such a trial design involving children are apparent.
Nonetheless, the Division would like to explore with applicants mechanisms for addressing such concerns and the possibility of studying infections such as uncomplicated urinary tract infections, uncomplicated skin and skin structure infections, uncomplicated abdominal infections, acute bacterial exacerbations of chronic bronchitis, acute or chronic sinusitis, and, perhaps, gonorrhea (although the public health implications of failed therapy here would need special attention). Perhaps these infections would be amenable to this approach, as therapeutic failure does not necessarily means acute life- or tissue-threatening consequences for the patient. Such trials would need to be conducted under tight clinical observation so that any therapeutic failure could be quickly noted and the patient begun on appropriate alternative therapy. Outside safety monitoring committees could perhaps also be utilized to assure that a study demonstrating a significant difference between treatment arms was stopped as soon as such was known without compromising the integrity of the trial. Depending on the number of arms selected and the therapeutic differences inherent in the different dosing regimens, the numbers of patients required in each arm could be considerably less that those required for routine active controlled trials. In any event, arguably, the data generated by such a trial could prove much more useful from a clinical and regulatory perspective than the data generated by a standard active-controlled trial trying to establish "equivalence". Economically, money spent by an applicant on each patient, would be spent garnering information on the applicants product, not re-establishing the effectiveness of a marketed competitive product.
The Division believes an applicant could use in vitro microbiologic susceptibility data and human pharmacokinetic/-dynamic data to select the initial dosing regimen believed to be the "optimal" dose. Dosing regimens above and several regimens below that "optimal" dose could be selected to try to demonstrate the shape of the dose-response curve. For some infections, the applicant may discover that, because of the self-limited nature of the infection, all doses are on the upper plateau of the dose-response curve. Though not as "clean" as a significant dose-response curve, such could still lead to rational choices about dosing based on the toxicity profile and therapeutic/toxic index of the product under study. In these situations, time to resolution of symptoms and eradication of bacterial infection, as appropriate, may offer an approach to product a "cleaner" dose-response curve. Either way, the Division believes such an approach should fulfill regulatory and scientific requirements for adequate and well-controlled clinical trial design.
Specifics of trial design, including masking, prospective clinical and microbiologic entry criteria, evaluability criteria, stopping criteria, and statistical analysis, should be developed prospectively in concert between the applicant and the Division. Input on trial design, execution, and analysis from outside experts, including our Advisory Committee, would be most welcome.
Once an appropriate dosing regimen is selected based on this approach, a confirmatory clinical-use trial of that dosage regimen could be conducted, as necessary. Dose-response data, from which relationships to both effectiveness and safety can be derived, along with confirmatory clinical use and safety trial data as required, could provide the requisite information for approval of a range of doses for the treatment of a specific infection that encompass an appropriate benefit-to-risk ratio. If discussion at the Advisory Committee undergirds this approach, applicants are strongly encouraged to consider this alternative approach to antimicrobial drug development.
DAIDP's Points to Consider
March 1995 Addendum : Helicobacter pylori-Associated Peptic Ulcer Disease
Date created: November 26, 1996; last update: July 6, 2005 |
|
