|
|
|
|
Pharmacoepidemiology and Drug Safety 9:253-255 (2000)
ISPE Commentary
We need a Postmarketing Drug Development Process! R. C. Nelson, Principle Consultant, RCN Associates, Inc. During my more than twenty years with the United States Food and Drug Administration’s (US-FDA) new drug evaluation and then the postmarketing surveillance programs, I worked on many interesting issues. I was exposed to and developed an understanding of a wide range of complex public health, scientific, and regulatory issues. It was a career that I would not have traded for any other. However, one item that I never understood was why all aspects of postmarketing surveillance (PMS) were so “unevolved” and ‘second-rate”. In the U.S. the explosion in the science of clinical trials was catalyzed by the 1962 efficacy amendments to the Food, Drug and Cosmetics (FD&C) Act. As the science of clinical trials evolved and matured in the 1970’s and 1980’s their inherent rigor became focused and the fully documented new drug (NDA) or market application (MA) became the summary product of a detailed, rigorous and well-documented development process. Regulators developed standards for trial content and conduct such as good laboratory practices (GLP), good clinical practices (GCP) and good manufacturing practices (GMP) to assure quality in manufacturing, testing, all within the framework of the humane and ethical treatment of study subjects. All processes are documented and subject to audit. All of these processes have evolved and matured markedly over the past thirty years. The related sciences also evolved and matured in lock step. Today, all efforts conducted in that pre-marketing realm are done so under GLP, GCP, and cGMP. Compliance to these and other important standards are assessed throughout. All data are validated, as are the computer systems in which they reside. All protocols are rigorously designed and followed. All submitted studies are drafted in full versions with all data and case reports made accessible to the regulators. The recent International Conferences on Harmonization (ICH) have served to further refine and update these good practices. As a result, the process of new drug evaluation is a validated one of rigor and high quality. The regulatory philosophy worldwide is one of setting standards for practice, allowing the industry to conduct the research and non-research efforts according to those standards. The submitted results are assessed and validated by the regulatory authorities. The situation on the postmarketing side was quite different and a consistent and compatible philosophy remains elusive. Spontaneous reporting was started over thirty years ago as an efficient and effective sentinel method to detect new, rare and serious adverse effects from newly marketed drugs. Systems began in United Kingdom and United States, and were rapidly supplemented by the World Health Organization’s (WHO) collaborating center effort, and now exist in over 45 countries. The stated intent from the beginning was to supplement this early detection system with a network of epidemiological data sources to confirm and quantitate the signaled risk. Just like that which was occurring in the premarketing arena in the 1970’s, the idea was to build a tool box (a conceptual framework) and fill it with the specific tools required to build and maintain a valid process of data acquisition and assessment. Spontaneous reporting was to be the first tool. Over the years many attempts were made to utilize other strategies and utilize other forms of data to create additional tools. Early strategies ranged from active case-control surveillance to large cohort designs. Methods of observational research were applied to all types of secondary, usually manipulating billing data in linked databases, in an attempt to assess risk associated with drug products. Later approaches included the collection of dedicated drug safety data [e.g. UK’s Prescription Event Monitoring System (PEM) and the general Practitioners Research Database (GPRD)]. In this present millennium, the U.S. still does not have any form of primary or dedicated drug safety data from which to assess rates of known reactions in clinical practice. We still do not know the incidence of known adverse reactions in the different clinical practice populations! Somehow unlike the carefully developed process of premarketing drug development, the postmarketing phase of drug development evolved differently and never followed a clearly agreed to conceptual framework. It also functioned within a contrary, unclear and non-supportive regulatory philosophy. Therefore, the existing PMS tools remained primitive because what they were intended to “build” remained unclear and unsupported. While pharmacovigilance and pharmacoepidemiological efforts abound, they do not, in my opinion, follow a logical regulatory philosophy and are conducted without a conceptual framework within which valid processes could be developed, standardized and adhered to. It is an “ad hoc” environment! As a result, today, the pharmaceutical industry expends an enormous amount of resources handling and reporting huge numbers spontaneous reports. These numbers are far, far greater than that which are needed to perform the primary function that this tool was developed to perform. Most of the additional data are not only unneeded but are subject to misuse. Rather than there being under-reporting of adverse drug reactions (ADRs) as many allege, I contend that there is more than a ten-fold over-reporting to these spontaneous systems. My judgmental estimates of industry-wide expenditures for handling spontaneous reports run 6000 persons and $1Billion USD annually. Most, if not all, of this effort is focused on solely being compliant with the variety of yet unharmonized regulations. Remarkably, only a few firms have sophisticated methods and processes in place to assess their subset of important spontaneous reports properly, alone or in any context. Quality varies tremendously. Some firms diligently accept what is offered by the reporter and follow-up in manner that is intended to receive no response, i.e. a 20-page general questionnaire. A few others rigorously query with specific related questions. Corporate performance metrics are focused on the number of reports processed and submitted to regulatory authorities, as well as the lack of formal regulatory warnings. Pharmacoepidemiologic studies, if thought of at all, is a tool reserved for emergencies. If a regulator pressures, some firms will pull together ‘experts with linked databases’ to study the issue. In addition to this dangerous lack of forward planning, pharmacoepidemiologic studies differ substantially from clinical trials in the degree of documentation submitted to a regulatory authority. Clinical trial submissions contain all case report forms, raw data, and full statistical methods. PE studies are usually in the form of a manuscript, pre-or-post publication. The International Society of Pharmacoepidemiology’s (ISPE) “Guidelines for Good Epidemiology Practices for Drug, Device, and Vaccine Research in the United States” were written to address this later deficiency. The effort was begun in response to challenge by the US-FDA to upgrade the quality of the work done in this field and to more clearly demonstrate that enhanced quality by properly documenting all conducted research. Though regulatory bodies have not officially yet accepted and adopted these guidelines, the industry need not and should not wait for such a formal action before they assure the quality of their observational research. Many years ago, the U.S. automobile maker, Ford, adopted “Quality is job one!” as their slogan. ISPE needs to consider a slogan with the same meaning. All PMS activities from the acquisition of spontaneous reports to the conduct of a prospective cohort study require a high level of scientific rigor. ISPE also must work with regulatory authorities and the pharmaceutical industry to develop a correct regulatory philosophy and an agreed to conceptual framework for data acquisition and assessment in the post-marketing phase of new drug development. ISPE must also advise on the goals of PMS. The ICH-E2C standards provide guidelines for a more comprehensive periodic report – the Periodic Safety Update Report (PSUR). The PSUR is a regulatory submission to the manufacturing application and is a continuation of the new drug development scientific and administrative processes. Its composition needs to be guided by “Good PMS Practices” that should not be second rate. The tools of PMS must be applied in a logical manner, be based on solid, validated, hopefully dedicated data, assessed rigorously and submitted in a full and documented manner, that is, within a agreed to Postmarketing Drug Development Process (PMDDP). As importantly, what are the goals of PMS or the PMDDP? The premarketing new drug development (NDD) has the goals of efficacy (usually two adequate and well-controlled clinical trials) and safety relative to intended benefit. Should the PMDDP goals be the early detection of new, serious reactions and the quantification of rates of reactions in the varied models of clinical practice within the first two years of marketing? Without clear goals there can be no framework. Without a framework there can be no agreed to process. Without a process, there can be no reasonable likelihood of success. Without a chance of success, there can be no real and sustained progress. PMS will remain subject to cyclical funding and the recipient of whatever priority is given by the latest crisis situation. I understand that the intrinsic rigor of a randomized clinical trial when compared to the potential rigor of an observational study, is one of the items that do need to differ. However, I contend that all else should be of the same high quality. If the current expenditures that I estimated above are anywhere close to being accurate, enough resources are already being expended to do PMS right! These resources need to be focused. The evolution of PMS must be redirected. I call on the PMS professionals to work with ISPE to decide on the goals of PMS. Create the blueprint, develop the tools, follow the work plan, and build the “safety house”, according to accepted ‘code’. Pharmacoepidemiology and Drug Safety 9:253-255 (2000) |
|
Last modified: 10/24/03 |