Risk Management Post Marketing

A Summary of FDA’s Guidance for Industry: Good Pharmacovigilance Practices and Pharmacoepidemiologic Assessment

Congratulations! Your company’s New Drug Application Department of Health and Human Services, Food and Drug (NDA) was approved and your drug marketing plan is a huge suc- Administration, and Center for Drug Evaluation and Research cess. Having successfully managed risk prior to approval and im- (collectively referred to as “FDA”) published in March 2005 a mediately thereafter, what steps should your company now take to Guidance for Industry: Good Pharmacovigilance Practices and Phar- assess and minimize safety risks? For many products, routine com- macoepidemiologic Assessment (Guidance). The Guidance applies pliance with postmarket requirements set forth in the Federal Food, to all drugs, except blood and blood components. While the rec- CDrug, and Cosmetic Act (FDCA) and Food and Drug Administra- ommendations set forth in the Guidance are nonbinding, they tion implementing regulations is sufficient for postmarketing risk represent FDA’s current thinking on what must be done to assess assessment. In some circumstances, however, unusual safety risks and manage risk post-marketing. More specifically, the Guidance may suggest a need for a formal pharmacovigilance plan. sets forth recommendations on (1) safety signal identification; (2) To help sponsors understand the importance of pharmacovig- pharmacoepidemiologic assessment; and (3) pharmacovigilance ilance activities in the post-approval period, the United States plan development.

What is Pharmacovigilance? compared to what would be expected to be plete and accurate to facilitate a meaningful associated with a product’s use.”2 The identi- evaluation of the relationship between the Pharmacovigilance means “all scientific fication of a safety signal generally triggers a product and adverse events. To help ensure and data gathering activities relating to the need for further investigation to determine the quality and usefulness of the reports, detection, assessment, and understanding of whether the product caused the adverse FDA encourages sponsors to use trained adverse events.”1 Good pharmacovigilance event in question. healthcare practitioners to query reporters of practice not only seeks to identify adverse adverse events. events but also attempts to provide an un- Identifying and Reporting According to FDA, good case reports in- derstanding of the nature, frequency, and Safety Signals clude the following elements: potential risk factors of these events. A spon- The first step in determining whether a • Description of the adverse event or dis- sor may accomplish this in large part by product caused a particular adverse event is ease experience identifying and evaluating safety signals. acquiring complete data from spontaneous • Suspected and concomitant product FDA uses the term “safety signal” to refer to adverse event reports, commonly referred to therapy details, including over-the-coun- “a concern about an excess of adverse events as “case reports.” These reports must be com- ter medications, dietary supplements,

6 Pro Te: Solutio Pro Te: Solutio 7 and recently discontinued medications patients, especially those with complicated Safety signals may also be evaluated • Patient characteristics medical conditions. through carefully designed non-randomized • Documentation of the diagnosis of the observational studies of the product’s use in event Methods of Investigation the “real world.” The Guidance focuses on • Clinical course of the event and patient There are a variety of methods for investi- three-types of non-randomized studies: (1) outcome (e.g., hospitalization or death) gating safety signals. One method recom- pharmacoepidemiologic studies, (2) regis- • Relevant therapeutic measures and labo- mended by FDA is the case review approach. tries, and (3) surveys. ratory data throughout event For this approach, FDA recommends that • Information about response to dechal- Pharmacoepidemiologic Studies lenge and rechallenge3 Pharmacoepidemiologic studies come in In some cases, adverse events are associated many forms. They may be cohort studies, with medication errors. Case reports involv- case-control, case-crossover, or others. Un- ing such errors should also include informa- like a case series, these studies employ strict tion about the product, the type of error, the protocol, utilize control groups, and test pre- work environment, type of personnel in- Good pharmacovigilance specified hypotheses. Pharmacoepidemio- volved, and contributing factors. practice not only seeks to logic studies may allow a sponsor to estimate

identify adverse events but the relative risk of an outcome associated Investigating Safety Signals with a product. A protocol for pharmacoepi- also attempts to provide an The Initial Investigation demiologic studies usually consists of clearly understanding of the nature, If one or more cases suggest that a safety specified objectives, a critical review of liter- signal warrants additional investigation, frequency, and potential ature, and a detailed description of the re- 4 sponsors should respond appropriately. This risk factors of these events. search methods used. Because pharmacoepidemiologic studies does not mean that sponsors should immedi- A sponsor may accomplish this are observational in nature, FDA recognizes ately deploy all available resources at the first in large part by identifying sign of a safety signal. The intensity and that they may be subject to confounding and and evaluating safety signals. method of investigation should be deter- bias, which make results of these studies more mined by the seriousness of the event report- difficult to interpret than other types of ed and by other factors, such as the report’s studies. Thus, investigators should seek to origin. FDA recommends that sponsors place sponsors assemble a case series and summa- minimize bias and account for possible con- an emphasis on reviewing serious, unlabeled rize descriptive clinical information to char- founding. One way to account for confound- adverse events, although other events may acterize the potential risk and, if possible, to ing is to conduct more than one study in warrant investigation. identify risk factors. A case series generally more than one environment. It may also be FDA also suggests that sponsors initially includes an analysis of several factors, such as helpful to use different designs. If these steps evaluate a safety signal generated from post- clinical and laboratory manifestations and are taken, consistent results may be evidence marketing event reports by reviewing indi- course of the event, demographic informa- that the observed results are accurate. Spon- vidual cases and then conducting a search for tion, duration of exposure, and other infor- sors are encouraged to communicate with additional cases. A sponsor may find addi- mation that may be useful in assessing risk FDA when pharmacoepidemiologic studies tional reported cases by searching the spon- and evaluating causality. are being developed. sor’s own databases, FDA’s Adverse Event Sponsors may also use statistical or math- Reporting System (AERS), or other available ematical tools, or so-called data-mining, to Registries databases. After gathering the necessary in- obtain additional information about the exis- The term “registry,” as used in the Guid- formation and reviewing cases, sponsors tence of an excess of adverse events reported ance, means: should look for features that may suggest a for a particular product. Because FDA recog- ...an organized system for the collection, storage, causal relationship between the use of a prod- nizes that statistical information does not retrieval, analysis, and dissemination of infor- uct and an adverse event. In making a deter- provide conclusive answers about whether a mation on individual persons exposed to a mination as to which cases suggest a causal product caused an adverse event, data min- specific medical intervention who have either a relationship, FDA recommends that sponsors ing techniques are not required. Nonetheless, particular disease, a condition (e.g., risk factor) not routinely exclude confounded cases (i.e., statistics may provide some insight into pat- that predisposes [them] to the occurrence of a cases with adverse events that have possible terns of adverse events reported for a given health-related event, or prior exposure to sub- etiologies other than the product at issue). product relative to other products in the stances (or circumstances) known or suspected Confounded cases are common among same class or to all other products. to cause adverse health effects.5

188 P Proro T eT:e Solutio: Solutio Registries allow sponsors to evaluate safety fied at-risk patient populations. Also, if ap- • The likelihood an adverse event repre- signals identified from spontaneous case re- propriate, the submission should propose sents a potential safety risk ports, literature reports, or other sources. They steps to further investigate the signal through • The frequency with which the event also facilitate the evaluation of factors that additional studies and propose risk minimi- occurs affect the risk of adverse outcomes, such as zation actions. Once the information is sub- • The severity of the event dose, timing of exposure, or patient character- mitted, FDA will make its own assessment of • The nature of the population(s) at risk istics. Whenever possible, FDA recommends the potential safety risk posed by the safety • The range of patients for which the that sponsors use a control or comparison signal in question. product is indicated group (i.e., individuals with a disease or risk • The method by which the product is factor who are not treated or are exposed to dispensed8 medical interventions other than the interven- FDA believes that pharmacovigilance plans tion of interest). Establishing a registry may may be appropriate where an analysis of these not be appropriate in all circumstances. To factors reveals the existence of serious safety determine whether to establish a registry, risks or that at-risk populations have not FDA recommends that sponsors consider the In cases where unusual been adequately studied. Sponsors may dis- types of additional risk information desired, safety risks become evident, cuss their safety concerns about a particular the attainability of that information through product with FDA and may also seek guid- a pharmacovigilance plan may other methods, as well as the feasibility of ance from FDA regarding whether a pharma- establishing a registry. Like pharmacoepide- be appropriate. A “pharmacovigi- covigilance plan is appropriate. miologic studies, registries should use written, lance plan” is a plan developed well-developed protocol. by a sponsor that is focused Conclusion Good pharmacovigilance practices and Surveys on detecting new safety risks pharmacoepidemiologic assessment are es- and/or evaluating already Surveys obtained from patients or health- sential elements of any risk management care providers may be useful in gathering identified safety risks. Such plan. Following FDA’s recommendations information about a number of things, in- a plan goes beyond routine will help sponsors obtain complete and ac- cluding safety signals and patient or provider curate information about safety signals, de- compliance and is designed knowledge about labeled adverse events. Like termine whether signals are indicative of pharmacoepidemiologic studies and regis- to enhance and expedite a safety risks, and then take steps to minimize tries, surveys should be developed using writ- sponsor’s acquisition of risk. Because it is impossible to identify all ten protocol. Protocol for surveys should safety information. safety concerns during clinical trials, it is provide objectives for the survey and a de- critical that sponsors use postmarketing tailed description of the research methods, safety data collection and risk assessment including: (1) patient or provider recruit- techniques to evaluate and characterize a ment and follow-up; (2) projected sample Do you need a Pharmacovigilance product’s risk and make informed decisions Plan? size; and (3) methods for data collection, to minimize such risk. 6 management, and analysis. Sponsors are As previously noted, in most cases, rou- encouraged to discuss their survey develop- tine compliance with FDCA requirements 1 Guidance, at 4. ment plans with FDA. and FDA regulations is sufficient to assess 2 Id. and minimize safety risks. In cases where un- 3 Guidance, at 5. What should you do with all usual safety risks become evident, however, a 4 Guidance, at 14. the information? 5 pharmacovigilance plan may be appropriate. Guidance, at 15. If, after conducting an investigation, a A “pharmacovigilance plan” is a plan devel- 6 Guidance, at 16. 7 sponsor determines that a safety signal may oped by a sponsor that is focused on detect- Guidance, at 19. 8 represent a safety risk, FDA recommends ing new safety risks and/or evaluating already Id. that the sponsor submit a synthesis of all identified safety risks.7 Such a plan goes be- available safety information and analyses per- yond routine compliance and is designed to formed to FDA. The submission should in- enhance and expedite a sponsor’s acquisition clude, among other things, an assessment of of safety information. Before developing a the benefit-risk balance of the product for the pharmacovigilance plan, a sponsor should population of users as a whole and for identi- consider several factors, including: Written by Adam Spicer

Pro Te: Solutio 9