print the article
send to a friend

Post-Marketing Surveillance: Better Late than Never
By Angelo DePalma
Date: 2005-05-20
Merck’s late-2004 problems with Vioxx, as well as the second look that all COX-2 inhibitors are now receiving, have focused attention on the need for continuing clinical trials even after a drug has met FDA approval. In addition, the concept of “life cycle product management” is growing in popularity in many diverse lines of business, including pharmaceuticals. The upshot is that both formal, Phase IV clinical trials and post-marketing surveillance are now a necessary and expected part of many manufacturers’ business practices. Clinical research organizations (CROs), software vendors and others are scrambling for a bigger piece of the business.


      ETRIALS            KENDLE               COVANCE               DATATRAK
In reality, surveillance and Phase IV studies have long been part of pharmaceutical products’ post-approval life. FDA requires sponsors to report any serious side effects not listed on the label within 15 days of learning of such events, and submit quarterly safety reports for three years post-approval. Post-approval testing and reporting experts acknowledge that, as rigorous as Phase I-III studies may be, they identify the only most common adverse reactions in “typical” patient populations. “Phase III testing involves thousands of patients, but after approval the numbers increase exponentially,” says Melanie Bruno, PhD, VP for Regulatory Affairs at Kendle, a Cincinnati, OH-based CRO. Higher patient counts offer the opportunity to observe rare side effects that went unnoticed during Phase III.
     Pre-approval clinical trials are neither large enough nor of long enough duration to provide everything anyone would ever wish to know about a drug’s safety. Adverse drug reactions, the impact of co-morbidities, inter-patient genetic variations, and drug-drug or drug-food interactions often go unnoticed during clinical trials and may only emerge post-marketing. Thus, initial safety warnings should be considered a work in progress, since they often change as Phase IV data rolls in. Phase IV surveillance may affect a drug’s labeling, how physicians prescribe it, or how companies market the drug and educate doctors and patients about it.
Simple to complex
Post-approval investigations vary widely complexity, from simple observational studies to double-blind, placebo-controlled trials. In addition to collecting drug safety data at various levels of sophistication, companies conduct formal studies to obtain specific label changes, for evaluating new patient populations, to monitor patient compliance and/or physician utilization, to uncover unusual pharmacokinetics, assess quality of life, or to determine optimal dosing levels or treatment regimens. Additionally, companies may decide to conduct follow-on animal studies or rigorous testing in specific co-morbid patients (particularly those with major organ or systemic disease) or the young.
     Sponsors obtain new post-approval indications through Phase IIIb studies, which are nearly identical to pre-approval Phase III trials. The differences between IIIb and IV are significant.
     Where FDA is a full party to Phase IIIB study results, the agency may or may not be a “customer” for Phase IV results. “Think of Phase IV as clinic-driven lifecycle management,” says Cynthia Verst-Brasch, PharmD, VP of late phase for Kendle. 
     Most Phase IV trials are conducted less formally (by far) than Phase III or IIIb trials, says Annette Stemhagen, Dr.PH, VP for Periapproval Services at Covance (Princeton, NJ). “Phase IV studies often need to be much broader [than Phase III trials],” she says. Phase IV investigations are conducted under actual-use conditions and tend to be observational. Otherwise, because of the sheer numbers of “subjects,” Phase IV would be far more costly and time-consuming than Phase III.
     “The success of a drug is basically determined by the marketplace,” says Dr. Stemhagen. “However, in the post-approval setting, many things Ð spontaneous adverse events, results of the sponsor’s committed post-approval studies, competitor study results pre- and post-approval, and others may hold serious implications to a drug’s market position and viability.”
 Dr. Stemhagen cites a recent example of a post-approval study for an asthma drug which identified a “safety signal” (unusual adverse event or event cluster) concentrated in a narrow patient population. Upon review, the sponsor decided not to remove the drug from the market but to address the problem with a labeling change that specifically mentioned
                                the at-risk patient sub-population.
"Think of Phase IV
as clinic-driven lifecycle
Phase IV studies
often need to be
much broader
than Phase III trials."

Committing to safety 

While safety has always been FDA’s top priority, the agency is increasingly requesting specific follow-up studies, even for over-the-counter products. Approximately three-fourths of recent New Chemical Entity approvals include at least one post-marketing commitment (PMC). Through PMCs, applicants agree to conduct post-approval studies on safety, efficacy, or use, or to address chemistry or manufacturing issues.
There are two types of PMC. A “Required” PMC is used as part of accelerated approval (for example deferred pediatric studies); an “Agreed Upon” PMC consists of studies undertaken to address a specific aspect of safety or efficacy. Examples include pharmacokinetic studies in subpopulations, long-term safety studies, and further “science” studies of the Phase I-II type.
PMCs tie in with industry’s and FDA’s new appreciation for risk management. Everyone realizes that if FDA demanded testing in every conceivable patient population no drug would ever be approved. Hence, FDA approves most medications based on an acceptable perceived risk-benefit ratio. PMCs are designed to shore up the “benefit” side of the quotient and minimize or at least identify the “risk” component, says Dr. Verst-Brasch of Kendle.
The goals of PMCs Ð maximum safety and efficacy Ð have been realized countless times, even for very well-studied drugs that went on to become blockbusters. In most cases the additional information benefited sponsor as well as patients. For example interferon alfa-2b (Schering-Plough’s Intron A), for treating hepatitis C, was originally approved for a six-month regimen but Phase IV studies showed that twice as many patients were cured if treatment lasted twelve months. Fluvoxamine (Solvay Pharmaceuticals’ Luvox), for obsessive-compulsive disorder, was approved in adults and appeared to work on pre-teens as well, but an additional pharmacokinetic study revealed that children in the 12-17 year group metabolized the drug more quickly than either adults or young children. This information led to higher dosing for teenagers.
Entering the 21st century
If problems with COX-2 inhibitors have taught us anything, it’s that Phase IV data require at least as much statistical validation as for Phases I-III. Unfortunately pharmaceuticals’ data capture capabilities are decades behind other industries’, says Jeffrey Green, Pharm.D., CEO of DataTrak International (Cleveland, OH). “Drug developers have implemented a range of high-throughput techniques on the discovery side, but they still pick up clinical data by hand and store it in three-ring binders,” he says. In the age of omnipresent computers and Internet, trial monitors still fly from study center to study center, correcting paper reports which are then carried or shipped by courier to central locations, where data are keypunched Ð twice for good measure. “The pharmaceutical industry tracks prescriptions every week but may wait for clinical data for many weeks, or months,” says Dr. Green, noting that post-marketing studies for Vioxx were done on paper.
DataTrak is one of several electronic data capture (EDC) firms that offer fully electronic, real-time data management for Phases I through IV and beyond. EDC instantly resolves data-related issues that raise safety or efficacy queries (but not, of course, the queries themselves). Individuals or groups authorized to view EDC data may view the product’s status in real time. Every piece of data is automatically audit-trailed and conforms to SAS data standards.
EDC provides a greater scientific basis for data collection, for example the opportunity to analyze data in ways that might not have been considered at the beginning of the study, or that might be extremely difficult with paper form-based data collection. From a legal/regulatory standpoint, real-time surveillance could exonerate companies posed with the question: “When did you know?” (This benefit works both ways, of course).
To avoid repeats of the COX-2 inhibitor recall fiascos, Dr. Green proposes independent review of Phase IV data utilizing EDC. “If FDA asks for Phase IV studies because of uncertainty regarding expanded exposure, the ongoing data collection should be reviewed by an independent committee with access to the Phase IV database as it is generated,” he says. “The data should be available in SAS data sets, so when they see something that approaches statistical significance, they can act. Otherwise, you have the situation where companies with huge financial interests are asked to make a decision that may be unfavorable.” Such a system may be deployed “today,” according to Green.
In Jan, 2005, HHS secretary Tommy Thompson promulgated a set of five initiatives, “Moving Medical Innovations Forward,” one of which was a mandate to FDA to standardize clinical trials based on EDC within the SAS data set framework. “EDC is even more important in Phase IV than in pre-approval trials because Phase IV entails more data and requires more rigorous data analysis,” says John Cline, CEO of etrials (Morrisville, NC). “FDA now asks some sponsors to follow patients for 15 years. How can you do that effectively with paper forms?”
For example, a CRO client of etrials recently was asked by European regulators for a statistical analysis that neither the sponsor nor the CRO had anticipated. Since the trial data had been collected on paper the CRO found itself in the unenviable position of inputting data from 8,000 patients, which took six weeks and barely met the timelines imposed by regulators. “Had they used EDC from the beginning, the data would have been accessible immediately and the job would have taken a few days,” says Mr. Cline.
EDC is not cost-justified for Phase I trials since building the back-end database and deploying the application is expensive and time-consuming. Sponsors and CROs really begin to see the benefits longer studies involving thousands of patients at multiple sites, studied over extended periods. In other words some Phase II studies but mostly Phases III, IIIb, and IV.
Current Issue

Continuing medical education (CME), as well as non-CME, promotional medical education (med ed), underwent dramatic change during the past year...