Medicines regulation WHO Drug Information Vol. 31, No. 2, 2017

Placebo and drug kits in clinical trial design

New and improved medicinal products are continuously needed throughout the world to prevent and treat diseases. Good quality clinical trials are key in bringing new safe and effective medicines to patients. Some information is outlined below on specific aspects of conducting clinical studies, namely the use of placebo as a control intervention and the use of drug kits for effective blinding of trials.

Introduction The use of placebo Before novel medicinal products are A placebo has been defined as “an inert introduced into widespread use, they must substance or sham procedure that is be assessed in clinical trials. Randomized provided to research participants with controlled trials (RCTs) are often considered the aim of making it impossible for them, the gold standard in this regard, although and usually the researchers themselves, to other study designs can also yield valid know who is receiving an active or inactive research results. Clinical trials should be intervention.”(1) designed in such a way that the effects of Placebos typically consist of the the experimental intervention are compared ingredients employed in the medicinal with those of a control intervention. In a product under study minus the active controlled trial, the subjects in the study and ingredient, making them inert. The inactive control group should be drawn from the ingredients (excipients) employed in a same population, and should preferably be pharmaceutical product must be “generally assigned to the groups by randomization to recognized as safe”1 for use in humans, remove bias in the allocation of participants. otherwise a medicinal product would not be Where feasible, clinical trials should be authorized for use. blinded, so that the subjects – and in In vaccine research, the term “placebo” is double-blinded studies also the researchers also applied to non-inert substances. In this – are unaware of who is receiving which context, an existing vaccine not studied in intervention. This helps to avoid behaviour the trial is added to both the investigational changes that may influence the study and the control product in order to avoid outcomes.(1) giving an “empty” injection to the subjects Two questions are discussed below in the control group. A disadvantage of that commonly arise in developing and this approach is that it complicates the evaluating clinical trial designs, namely in what situations it is acceptable to use placebo 1 The U.S. Food, Drug, and Cosmetic Act makes in the control arm, and how to achieve provision for food additives to be shown to be effective blinding. “Generally Recognized As Safe” (GRAS) through scientific procedures. Similar approaches are used in other jurisdictions.

This article originates from a request for advice sent to the WHO Prequalification Team. We thank Dr Matthias Stahl for his technical input. The article outlines selected concepts as found in published literature. It should not be considered as WHO guidance on the subject.

158 WHO Drug Information Vol. 31, No. 2, 2017 Medicines regulation

Placebo and drug kits in clinical trial design evaluation of the safety and reactogenicity of in the target population, (3) has been proven the vaccine under study.(2) ineffective in the target population, (4) has As the risks associated with the placebo an unknown or uncertain public health product are typically very low or non- impact in the target population which is existent, the use of placebos is generally to be evaluated against a placebo; or (5) is uncontroversial where there is no not acceptable to target population (e.g. established effective intervention for the a vaccine containing porcine gelatine in issue being researched. Where an established populations that have religious restrictions effective intervention exists, on the other on the consumption of pork). However, the hand, the use of placebo often raises risks and benefits of conducting a placebo- controversy among members of research controlled design should always be weighed ethics committees (RECs), regulators, and against those of alternative trial designs such policy-makers. The CIOMS Guideline 5, as response-adaptive designs, observational Choice of control in clinical trials, advises as studies, or historical comparisons.(1,2) follows: The use of placebo may require risk “As a general rule, the research ethics mitigation even if no established effective committee must ensure that research intervention exists. For example, in the participants in the control group of a trial Ebola vaccine trial conducted in Guinea of a diagnostic, therapeutic, or preventive the use of a placebo or an unrelated vaccine intervention receive an established effective in the control group was deemed ethically intervention. unacceptable as it would leave vulnerable Placebo may be used as a comparator individuals unprotected against Ebola without providing the established effective virus disease when a potentially effective intervention to participants only if: investigational vaccine was available. • there are compelling scientific reasons for Instead, vaccination of control subjects using placebo; and was delayed by 21 days, the minimal delay • delaying or withholding the established that would enable researchers to determine effective intervention will result in no more vaccine efficiency.(3) than a minor increase above minimal risk to the participant and risks are minimized, The use of drug kits including through the use of effective Specifically for researching new mitigation procedures.” pharmaceuticals, clinical trials should Risks and benefits of other study interventions preferably have a randomized double-blind and procedures should be evaluated according design. Blinding can take place at several to the criteria set out in Guideline 4 – levels: it may apply to the researchers who Potential individual benefits and risks of assign subjects to groups, the subjects research.”(1) themselves, the health care workers who A WHO expert group has identified five take care of patients in a study, and the situations when placebos may be acceptable researchers who record and assess the in the context of vaccine trials despite outcomes.(4) the existence of an effective intervention, A method of blinding clinical trials namely when the existing vaccine: (1) is is the use of drug kits, in which the not affordable or not accessible to the target investigational or control product is population, (2) has not been proven effective packaged for distribution to investigational

159 Medicines regulation WHO Drug Information Vol. 31, No. 2, 2017

Placebo and drug kits in clinical trial design sites. Each kit is labelled with a neutral ID, a kit is assigned to the next patient. This is without indication of its content, enabling successful blinding at the expense of wasting investigators to administer study or control one kit per subject.(5) drug to subjects in a blinded manner An analogous approach to the “waste- without the assistance of an unblinded one-kit” method could be used to blind pharmacist.(5) Each kit contains sufficient a trial where the control subjects receive product for a extended period, typically two a delayed intervention, by scheduling or three weeks,(6) although the quantity will additional visits in both groups. In the Ebola vary depending on the study design. vaccine trial conducted in Guinea this was The key component to the successful use not possible due to operational challenges; of drug kits is the creation of a kit list, where instead, to reduce the risk of bias arising kit IDs are randomly assigned to kit types from behaviour changes that might follow (investigational or control). This kit list is vaccination, participants were informed used at the time of manufacture to label the that it is not known if the vaccine works, kits, during shipping to track the kits, and and that they must still take steps to avoid during the study to assign kits to patients.(6) infection.(3) Many trials involve multiple clinical In practice, the design of algorithms for sites. To avoid the need for an unblinded coding and supplying blinded drug kits pharmacist at each site, use is often made of will take into account a range of factors a randomization centre to randomize study specific to each study, such as any additional subjects and manage the supply of drug kits trial arms, kit inventory size, enrolment i.e., track drug kit inventory at each site, rate at the sites, and times between subject ship kits to sites, and assign kits to patients. enrolments in relation to shipment In this way, information on whether a kit time for replacement kits. Operational contains test or control product does not characteristics of different types of kit accompany the kit but is available from the lists(6) and supply methods(5) have been randomization centre.(6) discussed in published literature. A criterion The coding used in creating the kit list has also been proposed for evaluating the and the distribution patterns and resupply strength of blinding in a clinical trial, even methods employed are key factors in if the researcher has been unblinded to blinding a clinical trial. A good design the contents of one or more kits. This is of should achieve a balance between kit interest because it is not uncommon for an efficiency and successful blinding. This investigator to be unblinded to a subject’s can be illustrated by two simple scenarios. treatment assignment for safety reasons or In a trial where each kit handed out to a from the subject’s adverse event or efficacy subject triggers a replacement by a single profile.(5) kit of the same type, no kits are wasted but the researcher can deduce that the patients Conclusions dosed with these two kits belong to the With today’s swift pace of product same study arm. In the opposite scenario development in a globalized market, the randomization centre would send designing, assessing, and authorizing clinical two replacement kits (one active and one trials can be challenging. Cooperation control) for each kit handed out to a patient, among regulators, ethics committees, and and would inactivate one of the two when sponsors to reach consensus on key ethical

160 WHO Drug Information Vol. 31, No. 2, 2017 Medicines regulation

Placebo and drug kits in clinical trial design and regulatory questions is essential, References and has proved particularly valuable in 1 International Ethical Guidelines for Health- situations of urgency and in low-resourced Related Research Involving Humans, Fourth Edition. Geneva: Council for International environments.(7) Organizations of Medical Sciences; 2016. Efficient conduct of the trials without 2 WHO. Expert Consultation on the Use of unnecessary regulatory barriers is equally Placebos in Vaccine Trials. Geneva; 2014. important. Excessively cumbersome 3 The ring vaccination trial: a novel cluster regulations, for example on importation randomised controlled trial design to evaluate and dispensing of placebos for clinical vaccine efficacy and effectiveness during trials, could delay access to effective and outbreaks, with special reference to Ebola. BMJ 2015;351:h3740. sometimes lifesaving medicines. WHO 4 Fathalla MF. A Practical Guide for Health stands ready to assist countries in identifying Researchers. Cairo: WHO Regional Office for well-balanced approaches for clinical the Eastern Mediterranean; 2004. trials, including those requiring the use of 5 Yu R, Coleman DA. Blinding Properties placebos. of Methods for Supplying Drug Kits to Lastly, well-designed clinical trials must Investigational Sites. Contemporary Clinical Trials Communications. 2015;1:22-27. be complemented by reliable post-approval safety assessment mechanisms. Many new 6 Coleman DA, Yu R. The Other Randomization – Methods for Labeling medicinal products are introduced early Drug Kits. Contemporary Clinical and/or exclusively into countries with Trials.2014;38(2):270-274. (Requires limited pharmacovigilance capacities. subscription) New guidance has become available on 7 The African Vaccine Regulatory Forum safety surveillance of vaccines, proposing a (AVAREF): A platform for collaboration in a public health emergency. WHO Drug structured process for evaluating whether Information 2015; 29(2):127-131. significant knowledge gaps exist, whether 8 CIOMS. Guide to Active Vaccine Safety passive safety surveillance is adequate, and Surveillance. Report of CIOMS Working if not, how active vaccine safety surveillance Group on Vaccine Safety. Geneva: Council studies can be designed and implemented.(8) for International Organizations of Medical Sciences (CIOMS); 2017. Available at: www.cioms.ch/shop. å

161