Post-Marketing Drug Safety Surveillance – Module 7, Session 7

>>William Douglas Figg:
The next lecture is provided by Dr. Timothy Jancel. Dr. Jancel received his Doctorate of Pharmacy
degree from the University of Pittsburgh, and a Masters in clinical research from Duke
University. He completed a pharmacy practice residency
at the University of Washington in Seattle, followed by a residency in infectious disease
at the University of California, San Francisco. Dr. Jancel is board certified pharmacotherapy
specialist, with the added qualifications in infectious disease. We hope you enjoy today’s lecture.>>Timothy Jancel:
Hi. Welcome, and good morning, afternoon, or evening,
depending on the time you’re watching this presentation. My name is Timothy Jancel, and today I’ll
be discussing introduction to post-marketing drug safety surveillance. In particular, I’ll be discussing pharmacovigilance
at the U.S. Food and Drug Administration and the Center of Drug Evaluation and Research,
otherwise known as CDER. Before I go on to the objectives, I just want
to go over our next two slides, which discuss all the acronyms and abbreviations we’ll be
discussing throughout this presentation. You can always refer back to these two slides
if you forget the abbreviation or acronym that we are using, since we do have a lot
throughout our presentation. Our objectives are located on this slide,
and by the end of the presentation, participants will be able to define pharmacovigilance;
describe the division of pharmacovigilance’s key safety roles in FDA’s Center for Drug
Evaluation and Research; understand components of post-marketing drug safety surveillance;
understand regulatory requirements and the role of MedWatch for reporting post-marketing
safety information; and finally, describe how adverse event reports are collected and
analyzed by the FDA, CDER, and the Division of Pharmacovigilance. To fulfil these objectives, I’ve organized
the talk into the following sections. First, I will discuss the FDA and where pharmacovigilance
fits in the big picture of drug regulation. Next, I will cover post-marketing surveillance
and spontaneous adverse event reports, and the FDA Adverse Event Reporting System, otherwise
known as FAERS. I will then cover signal detection. And finally, I will detail components of a
good case report, and also what makes a good case series, and the development and evaluation
of that case series. This slide briefly reviews the FDA’s overall
organizational structure, and the seven main centers that carry out the FDA’s public health
mission, which are divided by product type. This includes the centers for food, veterinary
medicine, medical devices, biologic products, drugs, tobacco, and regulatory affairs. We love acronyms at the FDA, and I’ve included
them for each center. The talk today will focus on drug safety activities
within the Center of Drug Evaluation and Research, or CDER. Looking closer at CDER, the drug safety activities
and surveillance discussed today take place in CDER’s Office of Surveillance and Epidemiology,
or we refer to as OSE. This slide displays the structure of OSE and
the main divisions and disciplines at work to fulfil our drug safety missions. In the bottom row, the divisions of pharmacovigilance
are highlighted. We’re divided into the Division of Pharmacovigilance
I and II. There are other divisions that include epidemiology,
or DEPI I and DEPI II, Medication Error Prevention, or DMEPA, and then, finally, risk management
— so, the Division of Risk Management, or DRISK. So, what is pharmacovigilance? Pharmacovigilance is defined by the World
Health Organization as the science and activities relating to the detection, assessment, understanding,
and prevention of adverse effects or any other drug-related problems. Examples of other drug-related problems include:
the wrong dose was administered; drug interactions; risk assessment; or name confusion or labelling
confusing. Based on the pharmacovigilance definition,
and our placement within the Office of Surveillance and Epidemiology, this slide details the major
responsibilities of the Division of Pharmacovigilance, which are the following: evaluating the safety
of drug and therapeutic biologic products; advancing public health by detecting and analyzing
safety signals; utilizing evidence-based methods; recommending appropriate regulatory actions,
including labelling changes, risk evaluation and mitigation strategies, or REMS, or other
necessary means. And finally, communicating relevant safety
information to the general public. We will go into detail into each of these
aspects throughout this presentation. The pharmacovigilance activities are carried
out by safety evaluators, or SEs, for short. They’re divided among 10 teams within Division
of Pharmacovigilance. Safety evaluators are typically clinical pharmacists,
and they provide clinic analyses of sources of post-market safety data to identify and
evaluate safety signals. Team coverage is based on relevant clinical
therapeutic area and is aligned within the Office of New Drugs and CDER, which is also
divided by therapeutic area. There are approximately four to seven safety
evaluators on each team, and each safety evaluator has a particular product assignment in their
respective clinical area. For example, my team covers the Division of
Transplant, Opthamalogic, Infectious Disease, including — we cover products that are antibacterials,
antivirals, which then align with the Division of Antiviral Products within CDER, as well
as the Division of Anti Infective Products in CDER. In addition to safety evaluators, the Division
of Pharmacovigilance has medical officers which provide expertise in various therapeutic
areas, such as dermatology, hematology, oncology, rheumatology, and pediatrics. These medical officers collaborate with pharmacovigilance
teams and the Office of New Drugs on safety evaluations. Now, I will discuss the science and principles
of post-marketing safety surveillance. Before we discuss post-marketing safety surveillance,
we need to discuss premarketing safety and safety in the overall life cycle of FDA-regulated
drug products. Many of you may have seen this graphic before. And not surprisingly, safety is addressed
in all aspects of the product life cycle, as you can see. This begins in preclinical, and continues
in Phase 1, Phase 2, Phase 3, and also post-marketing. At the bottom of the slide, it emphasizes
a critical part of the overall safety evaluation, whether prior to or following product approval,
is implementation of strategies and actions to minimize the risk of these identified safety
concerns. And the next few slides, I will go over some
of the difference between premarketing and post-marketing safety information. Although premarketing clinical trials are
the gold standard to determine safety and efficacy at the time of drug approval, many
limitations still exist. First, the size of the patient population
studied is small and only fairly common adverse events may be captured or reported — or reported. For example, drug-induced liver injury is
very serious adverse event, but it’s typically rare. Thus, if an adverse event occurs in every
10,000 patients, even if studies had thousands of patients in it, there is the potential
that drug-induced liver injury may never be captured. Also, a narrow patient population is usually
included. Thus, if certain drug interactions or adverse
events are more common in elderly, for example, they will not be captured in clinical trials,
since oftentimes this population is excluded. Also, children and pregnant women are rarely
included in these clinical trials Next, narrow indications are typically studied
and trials will often exclude patients with many comorbid conditions. Finally, trials typically have a short duration
and thus are not reflective of a drug’s product use and cannot identify safety concerns that
may occur after long-term use of the drug. In contrast, post-marketing safety monitoring
can address many of the limitations of premarketing studies, because it can better detect the
following. Low frequency reactions not identified in
clinical trials; these can be rare but serious events. Post-marketing monitoring can also address
high risk groups, like the elderly and children that aren’t typically included in clinical
trials. These data can also address long-term effects
like lipodystrophy with protease inhibitor used for the treatment of HIV. Drug-drug and drug-food interactions may also
be captured in post-marketing data. Finally, increased severity or reporting frequency
of known reactions can be monitored using post-marketing data, while clinical trials
typically only give a quick snapshot of the safety profile of the drug. Knowing the advantages of post-marketing monitoring,
here are various types of post-marketing surveillance data sources. First, spontaneous or voluntary reporting
of adverse events. These are cases that are found in national
sources, such as FDA MedWatch, or local or regional institutional levels, or in scientific
literature publications like case reports or meta analyses. Second, post-marketing studies. These studies can be voluntary or required
by the FDA. Types of studies include observational studies,
including those that are automated healthcare databases, or even randomized clinical studies. Finally, active surveillance sources, such
as the Drug-Induced Liver Injury Network, or DILIN, actively monitor cases of drug-induced
liver injury, as well as the FDA sentinel initiative, which uses information from healthcare
claims databases to identify or strengthen drug-related safety concerns. The remainder of the talk will focus on post-marketing
adverse event reporting. A key aspect of post-marketing safety surveillance
and how important you as a healthcare provider, consumer, or patient, are in the process of
pharmacovigilance. This slide describes how voluntary adverse
events reports are submitted to the FDA. Two pathways exist for patients, consumers,
and healthcare professionals to report a suspected adverse event. On the left-hand side: first, these reports
can be submitted through FDA’s MedWatch program, which encompasses a roughly 5 percent of all
reports. Or on the right-hand side, they can be submitted
by the product manufacturer, who are then required to submit all cases to the FDA. It is through this route that the vast majority
of cases are received and entered into the FDA Adverse Event Reporting System, or FAERS. It is also important to note that unlike some
other countries, the United States has no adverse event reporting requirement for practitioners. It is all voluntary. Note that for serious and unexpected adverse
events, the manufacturer is required by law to submit reports to the FDA within 15 days
of receipt of information. All other reports can be submitted periodically. It’s generally quarterly for the first three
years after approval, then annually thereafter, as per regulations. We will discuss the components of MedWatch
forms later. Under 21 CFR, or Code of Federal Regulations,
manufacturers must submit post-marketing safety reports to the FDA for the following: typically,
it’s either the 15-day alert report — these are serious and unexpected adverse experiences
from all sources. Or periodic adverse event reports — these
are domestic, spontaneous adverse events that are serious and unexpected, non-serious and
unexpected, non-serious and expected. And, again, it’s quarterly for the first years,
and then annually per guidelines. And this is taken verbatim from the 21 CFR. This defines what a serious adverse event
is per regulatory regulations. And a serious adverse event is any event that
results in death, life-threatening adverse experience, inpatient hospitalization — new
or prolonged, persistent, significant disability or incapacity, congenital birth defect, or
other serious. And this is based upon appropriate medical
judgment that may jeopardize the patient and require intervention to prevent a serious
outcome. Next, I’m going to detail how CDER and the
Division of Pharmacovigilance used the adverse event reporting system in FAERS to contribute
to our public health mission. As discussed earlier, spontaneous reports,
like those submitted to MedWatch and stored in FAERS, are one type of post-marketing report
used for safety surveillance. Spontaneous reports are communications from
an individual to a manufacturer or regulatory authority that describes a suspected adverse
event or events associated with a drug exposure. Spontaneous reports are passive and voluntary. They may be submitted by consumers or healthcare
professionals. The voluntary nature allows many factors to
affect the actual reporting. We’ll discuss some of the factors next. Here are some examples of factors that affect
spontaneous reporting, both from consumers and healthcare professionals. These include media attention. For example, in 2010, much publicity was devoted
to Avandia, or Rosiglitazone, and cardiovascular risk. Following discussions and media attention
of a FDA advisory committee, FDA saw an increased reporting of these events. Next is litigation. Many commercials advertise law offices to
help patients who may have suffered an adverse event; that can also stimulate reporting. Next, the seriousness of the adverse event,
the drug product, and the indication also can affect whether and how often an event
is reported. There are typically fewer reports for over-the-counter
medications. The number of reported adverse events for
a drug usually rises during the first few years after approval, and then typically declines
despite increases in prescribing rates. Next, the quality of the manufacturer’s surveillance
system may also impact reporting, as large companies have more resources to follow up
and engage with reporters, but smaller and newer companies may lack these resources. Finally, regulations affect reporting, particularly
in terms of what the manufacturers are required to submit to the FDA. The FDA Adverse Event Reporting System, or
FAERS, is a computerized database of spontaneous adverse events reports for human drug and
therapeutic biologic products. Collecting data since roughly 1968, over 13
million reports are currently stored in FAERS, with nearly 1.7 million reports submitted
in 2016 alone. This graph shows the number of adverse event
reports entered into FAERS by year. So, in the X axis, we have year, and in the
Y axis, we have number of reports submitted to FAERS. The purple and red colors represent reports
submitted from the manufacturer, while the light green represents the direct reports
from MedWatch. As I mentioned previously, roughly 5 percent
of reports are submitted directly to MedWatch. You can see the number of FAERS reports in
2016 is actually lower than 2015, because the FDA received a lot of non-expedited reports
in 2015 that were not from previous years. But overall, you can see a general trend that
the number of reports are increasing by year. FAERS is a drug safety surveillance tool,
and it has many strengths. It includes all U.S. marketed products, and
may include foreign products as well. FAERS includes all users, both approved indications
and off-label uses. For example, if a drug was approved for only
schizophrenia, we may receive adverse event reports for patients with bipolar disease
or depression. FAERS includes broad patient populations:
elderly, children, pregnant women, and patients with comorbidities who are often excluded
from clinical trials. FAERS is simple and relatively inexpensive
reporting system, as well. So, when is FAERS most useful? FAERS is ideal for events with small or rare
background rates. Examples include acute injury, serious skin
reactions, such as Stevens-Johnson syndrome, and is also useful for events that occur shortly
after exposure. So, what is the impact of having the database
of all these reports? FAERS allows us for the detection of events
not seen in clinical trials, and the identification of trends and reporting, possible risk factors,
such as certain patient populations, and other clinically significant safety concerns. FAERS is less useful for events with high
background rates, such as heart attacks and patients who have diabetes, since those patients
already have a high cardiovascular risk. FAERS is not helpful in the event — if the
event involves a worsening of a preexisting disease. The MedWatch form does not require certain
information, such as route of administration, product formulation, location of application,
genetics, or race. That’s if the issue is associated with one
of these characteristics, such as an issue with the oral solution, but it’s not occurring
with a tablet formulation, the FAERS database may not be useful. Comparing drugs, including those in the same
class, is difficult and often inappropriate. The time on the market and the actual drug
use might be different between products in the same class. This would be more appropriate for a clinical
trial or a post-marketing observational study with a controlled setting. If the intended therapeutic disease is reflected
in the adverse event report, it is also less useful. An example is the use of a psychiatric drug
for a psychiatric condition, and the adverse event is suicide. Finally, FAERS is less useful when looking
for drug interactions, and is subject to report biases and under-reporting. As discussed, FAERS is a large database with
many reports for all drugs and therapeutic biologics. It is often difficult and complex to find
what we are searching for. The next few slides will discuss how we are
able to manage these data and identify safety signals. And this picture just kind of shows what we
do at our job, and we have to find the signal, which is sometimes easy to miss. So, what exactly is a safety signal? There are many definitions listed on this
slide, and a true definition is really a hybrid of all these bullet points put together. Safety signals are reported information that
describe a possible causal relationship between an adverse event and a drug, the relationship
being previously unknown or incompletely understood. Examples of signals include new adverse events
or drug interactions, a new at-risk population, or greater severity or specificity about a
known event. We usually require more than one case report
to generate a signal. A great example of a safety signal includes
Raptiva, or Efalizumab, which was a treatment for psoriasis. During post-marketing safety surveillance,
progressive multifocal leukoencephalopathy, or PML, was discovered. PML was a rare and fatal demyelinating disease
that can cause brain lesions and can lead to death. PML is an opportunistic infection seen in
immunocompromised patients. In 2008 and 2009, three cases of PML in patients
receiving Efalizumab were identified in FAERS. In February 2009, the FDA issued a public
health advisory to notify healthcare professionals of three confirmed and one possible report
of PML. Then later, in April 2009, Genentech withdrew
Raptiva, or Efalizumab, from the market, consistent with FDA’s recommendations. Routine pharmacovigilance can include review
affairs, data-mining databases, which we’ll discuss more on the next slide, and periodic
safety reports from drug manufacturers. Other sources of safety signals may also be
included in routine pharmacovigilance and include the following: study reports from
clinical trials; epidemiologic studies or registries; case reports or studies in the
published medical literature; media; the New Drug Application safety database; and outside
inquiries, such as citizens’ petitions or interactions with foreign regulatory agencies. The list is endless and there is always new
resources, tools, and data streams to analyze. One tool that helps us identify safety signals
in the FAERS database is data mining. Data mining is a mathematical tool that involves
higher than expected reporting frequency of product event combinations. Data mining aids in finding reports and events
that are out of the ordinary. Data mining is intended to be hypothesis-generating
only, and may supplement the comprehensive review of the reports in FAERS. It does not confirm causality or replace the
clinical review of the case reports. Now I’ll discuss how to report in MedWatch. I discussed where safety information comes
from and what constitutes a safety signal. Now I will discuss how you can report an adverse
event to the FDA via the MedWatch system. And then, I’ll provide an example of a good
case report. The MedWatch program allows easy and convenient
reporting of suspected adverse events to the FDA. Here you can see MedWatch Form 3500 that is
a target for healthcare professionals. Many of you have probably seen or filled out
one of these — one of these forms. There are four official requirements for an
adverse event report to be accepted by the FDA. There has to be a patient identifier, a product,
an event, and a reporter. Additionally, there is a section for serious
outcomes, which are defined per regulations, and those include death, life threatening,
hospitalization, disability or permanent damage, congenital anomaly or birth defect, and other
serious. For devices, there is also a category of required
intervention to prevent permanent injury or damage. It’s not only voluntary to report in general,
but also voluntary on what you can report. Reports range from one-sentence descriptions
to an event, to carefully detailed narratives and accompanying laboratory details and medical
records. It all depends on how much information the
reporter is willing to share. Later on, we’ll discuss components of a good
report. In 2012, the FDA released a MedWatch form
specifically designed for consumers. This form, called MedWatch 3500B, is shown
on this slide. The form contains the same primary components
of the MedWatch Form 3500 on the previous slide; that includes a patient, a product,
an event, and a reporter. But it’s written at a reading level intended
for the general public. Input for this form came from consumer advocacy
groups, as well as the general public and — when they were developing this form. There are two ways to report MedWatch form,
you can do it online on the website listed, which is the easiest way. Or the form can be downloaded from the site,
completed, and then mailed or faxed to the FDA. Pharmacovigilance work at CDER depends a great
deal on FAERS cases. We receive about 1 million reports annually,
and as I’ve stated previously, these data are quite heterogeneous. The quality of the reports and the date in
FAERS are highly variable, yet the public health impact of signal detection and pharmacovigilance
is dependent on these data. Therefore, we want to present the components
of a good case report. I will begin with example case one. A healthcare worker reported a male patient
started Drug X at 5 milligrams daily for type 2 Diabetes on February 11th. On an unknown date, the patient developed
liver failure. Additional information was not reported. So, that was an example of a bad case report
that we would receive. Here’s an example of case two, or a good case
report — or best case representative. A 59-year-old male with type 2 diabetes, hyperlipidemia,
and hypertension, with no history of liver disease, started Drug X on February 11th. Other concomitant medications included Simvastatin
and Lisinopril. Baseline labs are also drawn on February 11th
and reveal liver enzymes, INR, creatinine, and bilirubin — all within normal limits. The patient also had no alcohol use. Eight weeks after starting Drug X, the patient
presented to the emergency room with a five-day history of jaundice, dark urine, and nausea
and vomiting. He was admitted to the intensive care unit,
and subsequently diagnosed with acute liver failure. Drug X was stopped upon admission. Viral hepatitis serologies ruled out any sort
of infectious etiology, and seven days after stopping the medication, all lab values returned
to normal. So, that was an example of a good case report. And so, the components of a good report are
listed here. Description of adverse event, suspected and
concomitant product therapy details. Ideally, we would like to know the dose, the
dates of therapy, any patient characteristics including sex, age, weight. Any baseline medical conditions, comorbid
conditions, family history, or other risk factors of interest. We want a documentation of the diagnosis,
the clinical course and outcomes, relevant therapeutic measures and laboratory data,
dechallenge and rechallenge information, reporter contact information so that we can follow
up for additional details, and any other relevant information. Now, we will discuss the development and evaluation
of a case series for an adverse event. For example, as we develop a case series — this
is just an example of what would occur at the FDA. A safety evaluator comes upon a MedWatch report
describing a patient taking a newly approved product for diabetes. The report describes a patient who developed
liver failure six weeks after starting the product. Using this example, we will walk through the
case series development and signal detection process. So, as a safety evaluator, we have identified
an event of interest. Then we use our knowledge of the clinical
course of the disease to formulate a case definition, which may include both clinical
features and laboratory findings. We conduct a search of the FAERS database
using MedDRA terms. MedDRA stands for medical dictionary for regulatory
activities. All FAERS cases are coded using MedDRA terms. Thus, for our possible liver failure signal
FAERS case, we will then query FAERS and look for additional cases with events related to
liver failure for the drug in question. This not only includes diagnoses or clinical
presentations, but also laboratory information. A thorough search of our example would include
terms such as acute hepatic failure, alanine aminotransferase increased, transaminases
increased, transaminases abnormal, encephalopathy, and sometimes even demographic information,
just to name a few. After a thorough search of the database, the
next step is to evaluate the cases to determine how many cases to include in our case series. The main principles are to determine the temporal
relationship and any causal association. Temporal relationship is important; if we
cannot be sure that the adverse event occurred after the drug was taken, perhaps we don’t
have a good case. Many times, we could receive a case, but they’ll
never state the date that the drug was started, so then we’ll have to exclude that case in
our case series. Causality assessments like the one developed
by the World Health Organization, classify cases in a range from unlikely to certain. Several factors determine the strength of
associations, and these factors may include the following: any dechallenge or rechallenge
information; did the event improve after stopping the drug? And/or did it recur if the drug was restarted? Comorbidities or concomitant medications,
could they have contributed to the development, or even caused the event? And there are data to support that such an
adverse event is biologically plausible based on the pharmacologic effects of the product. Of note, sometimes after reading through the
cases, new information or revelation may come up. So, a new case definition and a new search
may be needed. So, once our evaluation of the safety signal
is complete, what does the agency do with this information? And this slide just graphically shows all
the possibilities. But the next slide goes through the several
action items that are possible and the appropriate next steps based on the strength of the data
and the association, the severity and significance of the event, the predicted frequency, and
the overall health — public health impact. So, what changes could be made, or what regulatory
actions could be made? Product information changes could occur, adverse
events could be added to the warnings and precautions, or adverse reaction sections,
for example. Or an adverse reaction could be elevated within
the label from post-marketing experience to a warnings and precautions. New pharmacovigilance activities, such as
enhanced surveillance, include expedited reported, or creation of a registry, or there might
be a need to perform epidemiologic studies. A risk evaluation and mitigation strategies,
or REMS, can be created, or a preexisting one can be modified. Drug safety communications, or DSCs, are also
common and are an effective way of sharing and explaining this new safety information
to healthcare professionals and patients. And finally, market withdrawal is possible,
but typically only in extreme scenarios. So, now I’ll discuss communicating safety
issues. This slide contains examples of how new safety
information is shared with the public. This is for both emerging and established
safety concerns. MedWatch safety alerts: these are often — these
often provide links to drug safety communications. An example of a recent MedWatch alert and
DSC from 2013 detailed the risk of potential permanent nerve damage with the use of fluoroquinolones. FDA drug safety newsletter for health professionals
also provides drug safety-related news. In accordance with section 915 of the Food
and Drug Amendments Act of 2007, also known as FDAAA, the FDA website contains links to
post-market drug safety information to improve transparency and communication to patients
and healthcare providers. As part of Section 915, post-marketing evaluations
are performed when at least 18 months has passed after drug approval or its use has
reached at least 10,000 individuals. The findings of these reports are posted to
the public. Also, Section 921 mandates the posting of
quarterly reports of the affairs website of any new safety information, or potential — potential
signals or serious risk identified by FAERS within the last quarter. And all — everything that I’ve just discussed
with FDAAA will be ending at the — September of this year in 2017, and has been replaced
by something that will be coming with the passage of the 21st century Cures Act. Additionally, the pharmacovigilance work is
sometimes published in the medical literature or presented at scientific meetings. Finally, important safety information is also
shared with our foreign regulatory counterparts, with a clear public health objective. This screenshot of the MedWatch FDA internet
page is a one-stop shop for most of these examples on the previous slide. The new MedWatch alerts are seen at the top,
and then on the left, past alerts, MedWatch sign up, and the safety newsletter can be
found. And at the bottom are the FDAA 915, 921 postings
at the bottom. The web address for the one-stop shop is You can sign up to receive emails or tweets
regarding safety alerts for drugs, devices, vaccines, other biologics, dietary supplements,
and cosmetics from MedWatch. And then, finally, there’s a reference slide
to links to all of our — with everything we’ve discussed today. And I double-checked all these links; they
should be working as of today. And that ends our presentation, and thank
you very much for following us along on the discussion of pharmacovigilance at the Food
and Drug Administration.

Leave a Reply

Your email address will not be published. Required fields are marked *