In this report, an expert witness retained by Pliva expresses the opinion that Pliva's metoclopramide warning label was adequate.
9. However, because of how the FDCA was structured, companies that wanted to
market generic versions of drug products approved under NDAs after 1962, such as
Reglan brand of metoclopramide, would have had to duplicate the expensive animal
and clinical studies the innovator did, even though the patent protection had
ended. This situation had the consequence of providing a virtual monopoly for
drugs first approved after 1962, because generic drug manufacturers did not have
the financial resources to perform extensive clinical and animal studies and to
prepare and submit NDAs to the FDA. The Congress remedied this situation in 1984,
when the Hatch-Waxman amendments to the FDCA extended a simplified administrative
process the FDA used for generic versions of older drugs, called an Abbreviated
NDA (ANDA) to generic versions of drugs first approved after 1962. The public
policy arguments for the Hatch-Waxman amendments included the following:
United States District Court, W.D. Louisiana.
TARVER et al,
v.
WYETH INC et al.
No. 04CV02036.
February 1, 2007.
Expert Report of James Morrison
1. I have been retained by defendant Pliva, Inc., to render expert opinions in
this matter. I have been a consultant and expert in drug regulatory matters for
about four years. Before that I held a variety of positions with the U.S. Food and
Drug Administration (FDA) for thirty-seven years, including senior management
positions, all dealing with the regulation of drugs. My training and experience
are listed in my resume (attached as Appendix A), along with a list of all
publications that I have authored or co-authored within the preceding ten years
and a list of all cases in which I testified or was deposed in the past four
years. My fee for consultation, preparation, deposition and testimony is $440 per
hour or $4000 per day for work outside the Washington, DC/Baltimore, MD area
2. In addition to my resume, there is some relevant background experience that I
would like to note. As a compliance officer in the Bureau of Drugs, I made
recommendations and decisions regarding the new drug status of products, and
reviewed and recommended action on proposed legal actions against drug products,
firms and individuals on drug regulatory issues, including new drug and labeling
charges. As Director, Regulatory Affairs Staff in the FDA's Bureau of Drugs, I
actively participated in a rewrite of the new drug regulations, including those
cited in this report. As Deputy Director, Office of Drug Standards, I was the
FDA's primary contact and negotiator with the Congress in drafting the
Hatch-Waxman Amendments of 1984, which enabled the marketing of generic versions
of drugs first approved after 1962. I chaired the committee that drafted the FDA's
regulations implementing that law and edited those regulations. I also chaired a
committee charged with investigating reported therapeutic failures of generic drug
products. As Special Assistant to the Bureau Director and later as Ombudsman and
Senior Advisor to the Director of the Center for Drug Evaluation and Research
(CDER), I participated in and chaired committees charged with evaluating adverse
event reports, investigating and recommending action on various products that
posed potential health threats.
3. The attached Appendix B lists all documents I relied on in formulating my
opinions, in addition to my education, knowledge and experience. My opinions are
based on the materials I reviewed, and I reserve the right to amend them in light
of new information or the production of new documents and opinions.
SUMMARY OF OPINIONS
4. In this report, I opine that Pliva could not legally have changed the labeling
of its product to strengthen warnings about possible adverse events from using its
metoclopramide product. The FDA requires that labeling for drugs such as Pliva's
metoclopramide must follow exactly the labeling of Wyeth's Reglan metoclopramide
product.
5. I also opine that Pliva, as an ANDA holder, did not have the responsibility
under the Federal Food Drug and Cosmetic Act (FDCA) and FDA rules to monitor the
world literature or the U.S. literature on metoclopramide. I describe Pliva's
adverse event reporting and complaint handling obligations under the FDA's rules.
STATEMENT OF OPINIONS AND BASES
Organization and History of the FDA
6. The FDA has the primary responsibility for the regulation of drugs in this
country. The Center for Drug Evaluation and Research and Evaluation (CDER) is the
FDA's drug organization. Within CDER, there are offices that:
* Review and approve applications for marketing new drugs (NDAs)
* Review and approve applications for marketing generic drugs (ANDAs)
* Develop monographs for non-prescription or over-the-counter drugs
* Monitor and evaluate the safety of drugs already on the market
* Enforce the FDA's laws and regulations.
7. The FDA enforces the Federal Food Drug and Cosmetic Act (FDCA). The Act was
originally passed in 1938. It required drug companies to submit data to the FDA
regarding the safety of drug products they intended to market. If the companies
did not hear from the FDA in sixty days, they could proceed to market their
products.
8. In 1962 the FDCA was significantly amended, and drug companies were required to
submit new drug applications (NDAs) to the FDA that contained data to demonstrate
the safety and effectiveness of the proposed drug products, and they could not
market the products until the FDA affirmatively approved.
9. However, because of how the FDCA was structured, companies that wanted to
market generic versions of drug products approved under NDAs after 1962, such as
Reglan brand of metoclopramide, would have had to duplicate the expensive animal
and clinical studies the innovator did, even though the patent protection had
ended. This situation had the consequence of providing a virtual monopoly for
drugs first approved after 1962, because generic drug manufacturers did not have
the financial resources to perform extensive clinical and animal studies and to
prepare and submit NDAs to the FDA. The Congress remedied this situation in 1984,
when the Hatch-Waxman amendments to the FDCA extended a simplified administrative
process the FDA used for generic versions of older drugs, called an Abbreviated
NDA (ANDA) to generic versions of drugs first approved after 1962. The public
policy arguments for the Hatch-Waxman amendments included the following:
* Brand name drugs (those with approved NDAs) were so expensive that many patients
on fixed incomes could not afford them and so never filled their prescriptions.
* The FDA's new drug regulatory processes created monopolies outside the patent
system and unreasonably extended patent protection for some drugs.
* This extension of patent protection was not intended by Congress when it passed
the 1962 amendments to the FDCA.
* The potential savings to consumers from a competitive generic drug market were
substantial. In recent years, savings from the availability of generic drugs have
been estimated at around $10 Billion annually.
FDA's New Drug Approval Process
10. Section 505 of the FDCA requires that most prescription and new
nonprescription drug products must be approved by the FDA before they can be
marketed in the US. For active drug ingredients, combinations, routes of
administration or dosage forms that have not previously been approved, the FDA
requires that a NDA be submitted by the sponsor and be approved by the FDA before
they can be marketed. The sponsor of such a drug product must show that the
product is both safe and effective when used in accordance with its proposed
labeling. To demonstrate the new drug's safety and effectiveness, the sponsoring
drug company must conduct extensive studies on the drug using animals and humans
in sound, scientific studies. Such studies may cost tens or hundreds of millions
of dollars. After patent protections expire, generic drug manufacturers may submit
ANDAs for generic versions of drug products approved in NDAs. The ANDA must
include information about the bioequivalence of the product compared to the NDA
product and the chemistry, manufacturing controls and other aspects of the product
to assure therapeutic equivalence to the NDA product. However, no clinical safety
or effectiveness studies are required. In addition, the FDCA requires that the
labeling of the generic drug product copy exactly the content of the labeling of
the NDA holder's product to which it is equivalent. [21 U.S.C. 355(j)]
NDA Holders' Role in AER Evaluation and Labeled Warning Changes
11. For prescription drug products that are marketed under approved NDAs, the NDA
holders are regarded as experts in their drugs. The NDA holders have performed
years of testing of their drugs, both in animals and in humans. The companies have
staffs of clinicians, epidemiologists and other health professionals who can
analyze adverse event reports (AERs) and complaints, monitor the world literature
for updated information about their drugs, and prepare reports of their findings
for transmission to the FDA. Therefore, NDA holders are required to provide
periodic reports to the FDA of analyses of AERs they received and updates from the
world scientific literature relating to their drugs. In addition, they are
required to notify the FDA within fifteen days of any serious and unexpected
reports, either from AERs or from the literature. [See 21 CFR 314.80] They may
also propose changes to the labeling of their NDA products, but FDA approval of
the changes is required.
Generic Drug Manufacturers' Role in AERs, Labeling and Warnings
12. For prescription generic drug products marketed under approved ANDAs, firms
are required to submit AERs, fifteen day reports on serious and unexpected adverse
events and limited periodic reports to the FDA, but they are not required to
monitor the world scientific literature for information on adverse effects of the
drug or to analyze the literature and report findings to the FDA as NDA holders
are required to do. [See 21 CFR 314.98 and 314.80(b)] . Section 314.80(b) of the
FDA regulations begins with the phrase, "Each applicant having an approved
application under 314.50 ..." An application under 314.50 refers to a NDA, not an
ANDA. Thus, as interpreted by the FDA and in my experience at the FDA, both when I
headed the staff responsible for writing drug regulations and when I had
management oversight over the review of generic drugs, 314.80(b) applies only to
approved NDA holders. Pliva holds an approved ANDA for metoclopramide, and so is
exempt from the provision requiring NDA holders to analyze adverse event reports
and to monitor the world scientific literature for the approved drug. Because
generic drug manufacturers make many different drug products and because they do
not have the staffs of clinical and epidemiological experts that NDA holders
usually have, they are not able to perform extensive clinical and epidemiological
analyses on each drug product they manufacture. Adding such burdensome
requirements that would be duplicative of functions already being done by the NDA
holders and the FDA would undermine the cost saving role of generic drugs. Many
generic drug manufacturers market a broad array of products, and the economic
burden of duplicating the NDA holders' and the FDA's functions of epidemiological
surveillance of all drugs world-wide would add significantly to the cost of
generic drugs for little or no gain in safety to the patient. Generic drugs are
not introduced into the market until there is substantial experience with the
safety, effectiveness and adverse effects of the innovators' drug products. For
these reasons, the FDA's regulations on periodic reports are divided so that the
section dealing with requirements for monitoring the world literature and
clinically analyzing AERs is limited to NDA holders. The other sections of 314.80
apply to ANDA holders as well as NDA holders.
13. ANDA holders are required to have the same labeling for their generic products
as that of the NDA product. The ANDA holders cannot change their labeling unless
the referenced NDA product has changed its labeling. They and the NDA holders must
get FDA approval for any change they propose to make in drug product labeling.
[See 21 CFR 314.70(c)(6)] While ANDA holders may seek to have the FDA change the
labeling for all marketed products of the drug to reflect new information, it
would be extraordinary for a generic manufacturer to do so. I have not seen such
an action by a manufacturer for a generic drug in my career at the FDA. This makes
sense, because a generic drug manufacturer would need to provide extensive
information to the FDA to make a convincing argument that the labeling should be
changed. However, the generic company would not generally have such information,
is not required by the FDA to obtain such information, and does not have the
resources to amass such information while staying price competitive in the
marketplace.
FDA's View of Its Regulatory Responsibilities for Labeling
14. The FDA regards drug labeling as an essential part of the safe use of drugs,
and it does not permit companies to change the labeling of drug products without
its concurrence. There is a public health policy basis for the FDA's labeling
rules. If drug companies were given license to change product labeling in response
to isolated reports of adverse events associated with multi-source products,
serious confusion would result, and FDA's role as the primary drug regulator in
the US would be undermined. Further, if companies added warnings or otherwise made
labeling changes based on anecdotal information contained in AERs or literature
reports, the additional labeling statements, not based on sound scientific
information, would expand the length of drug labeling and dilute the effect of
warnings that are appropriately in the labeling. For generic drug products the FDA
is the organization with the facilities, staff and systems to monitor for and
detect new adverse drug experiences. The FDA can then determine through an open
process of notice and comment rulemaking and advisory committees whether labeling
changes are warranted. It is useful to discuss in greater detail the adverse event
monitoring system used by the FDA, the role of drug manufacturers in it, and the
problems associated with drawing conclusions from the data in the system.
The FDA's System for Monitoring Adverse Event Reports
15. The FDA maintains a system whereby consumers and health professionals may
report adverse events they have experienced or seen first-hand. The FDA receives
approximately 300,000 such reports annually. That system is called MedWatch, and
its database is called AERS. For drug and biological products, reports to the
system are voluntary. Although most major drug regulatory agencies around the
world use such systems, these spontaneous reporting systems have known
limitations, including those described below:
16. Variable reporting rates: Because reporting is voluntary, there are varying
degrees of reporting of adverse events. The reporting varies because of a number
of factors, including: whether the consumer or health professional notes an
association between the adverse event and the use of the product, and whether
there are disincentives or incentives for health professionals and consumers to
submit reports.
17. Whether an adverse event is recognized as associated with a product depends in
part on whether the health professional or consumer expects that such a reaction
would be likely to occur. That expectation is strongly influenced by history and
publicity. For example, the thalidomide disaster in Europe in the early 1960's was
caused by mothers taking a newly approved drug, thalidomide, which was used to
treat morning sickness. As a result, thousands of babies in Europe were born
missing limbs. When I investigated a potential causation of limb reduction birth
defects by the drug Bendectin, which was marketed in the 1970's for morning
sickness, I was told by experts at the Centers for Disease Control[FN1] that there
was a several-fold increased reporting rate of such defects associated with drugs
because of the past association between limb defects and thalidomide. This is
consistent with my training and experience in analyzing adverse event reporting.
Similarly, when there are news stories that associate deaths and serious injuries
with a particular product, there is likely to be an increase in reports to the FDA
of associations with similar products. The increase in reports is seen, regardless
of whether the media stories attributing a causal link between the product and the
death are accurate or not.
FN1. Conference call from Dr. I. Richard Crout and me at the FDA to Dr.
Oakley and others at CDC.
18. Lack of numerator and denominator: Another weakness of the spontaneous
reporting system is that the total number of actual adverse events associated with
a product is not known, nor is the number of marketed units of the product that
were used at the time of any given event known. This is especially true if sales
of the product are increasing or decreasing rapidly. Thus, the raw number of
adverse events reported for a product over time does not convey an accurate
picture of what is happening with the product. For example, if reports are
increasing over time, the increase might reflect increased sales of the product or
an increasing health problem, or neither.
19. Association is not the same as causation: Even if the number of units sold was
known accurately and the number of actual adverse events was also known, there
would still be the problem of determining whether the association is mere chance
or whether it really demonstrates causality.
20. Quality of data in the system: The AERS system suffers from a chronic problem
of incomplete data submitted with reports. Accurate information on concomitant
medication used by the patient, how much of the product was consumed and for how
long may be missing. Sometimes the reporting physician does not know the complete
information. When adverse event reports on drug products come to the FDA, the CDER
Drug Safety staff evaluates them as to their potential significance. If the report
involves serious, unexpected events, a staff member reviews the report to
determine if all needed data are included. If it is not complete, the staff member
contacts the reporter to try to obtain the missing data. Even the follow-up calls
may not elicit sufficiently complete information to make the report useful.
21. Because of these factors, spontaneous reporting systems such as FDA's MedWatch
are viewed primarily as tools to flag potential problems with drugs and other
medical products and to generate hypotheses for further study. I reviewed AERs
received by Pliva for 5 and 10 mg metoclopramide during a five-year period ending
in 2004, and I noted there were only seven reports involving extrapyramidal
symptoms, two of which were tardive dyskinesia. For the reasons I discussed above,
it is impossible to draw definitive conclusions from such data. However, if
metoclopramide labeling underestimated such adverse events by a factor of one
hundred, as plaintiffs suggest, one would expect many more AERs and, more
importantly, much more attention to the problem in the medical literature than has
been cited in plaintiffs' allegations.
FDA Studies into Metoclopramide Use
22. The FDA assumes a role in the evaluation of AERs independent from the role of
NDA holders. The FDA's Office of Drug Safety in CDER uses AER data coupled with
information on drug use obtained from private contracts to add to the available
literature on drug adverse effects. Two publications that resulted from such
evaluations are of particular interest here. In 2004, FDA staff published an
article[FN2] that demonstrated that use of metoclopramide declined following the
1993 introduction of cisapride, another drug for treating gastroesophageal reflux
disease (GERD). However, they reported that metoclopramide use rose to earlier
levels and continued to grow after the 2000 withdrawal of cisapride due to serious
cardiac adverse events. In addition, the authors reported on their retrospective
analysis of 87 cases of tardive dyskinesia (TD) AERs in the FDA's MedWatch
database. These reports were associated with long use (mean > 2 years), along with
other TD risk factors, including age, the concomitant use of antipsychotic and
other drugs known to increase the risk of TD. The authors urged caution in the use
of metoclopramide in patients at risk for TD or in prolonged use of the drug.
Although they cited as references articles by Ganzini and Sewell, who estimated
the risk of TD as much higher than current metoclopramide labeling, they did not
recommend changing the labeling of metoclopramide, nor did they question that TD
was a rare side effect of the drug.
FN2. Shaffer, D, et al, Tardive Dyskinesia and Metoclopramide Use: Effects
of Cisapride Market Withdrawal, J. Am. Pharm. Assoc. 2004; 44(6); 661-5.
23. In 2006, staff from FDA's Drug Safety Office published an abstract, Adherence
to Metoclopramide Duration of Use Recommendation: Claims Data Study, in the 12th
Annual FDA Science Forum (April 2006) based on their work done in preparation for
FDA's review of a pending NDA for a product containing metoclopramide and
naproxen. In the abstract, they reported that, based on three years of claims from
Caremark (200,907 patients with 572,205 prescription claims for metoclopramide),
cumulative therapy for longer than six months was found for more than ten percent
of patients. However, for eighty-nine percent of patients, the duration of therapy
with metoclopramide was no longer than the twelve week limit recommended in the
labeling.
Specific Comments Related to This and Similar Cases
24. There are a number of statements in both the report and deposition of
Plaintiffs' expert, Bob West, PhD., with which I disagree. However, rather than
addressing each of them, the following represent major areas of concern.
* On page 8. of his report dated August 2006, he says that, "The indication and
warning to avoid long term use should have been placed in a single distinctive
black box at the top of the label. Responsible manufacturers use this technique to
highlight important warnings for physicians." However, black box warnings are a
specific class of warnings generally regarded as the most serious type of warning
that can be placed on a drug product. Drug companies do not generally propose such
warnings, because it is a signal to physicians that the product is among the most
dangerous of all marketed prescription drugs. The FDA does not suggest or require
such warnings lightly, because it reserves black box warnings for those drugs with
extremely severe, often fatal adverse effects. In addition, placing such a warning
on a drug's labeling restricts how it can be advertised. Thus, Bob West
trivialized the black box warning in describing it as. used by responsible
manufacturers to highlight important warnings. The FDA would not, in my opinion,
agree that metoclopramide should have a black box warning, and it would not
approve any NDA labeling supplement to that effect.
* Bob West goes on to state on page 10 of his report that FDA regulations allow
approved application holders "to add or strengthen a contraindication, warnings,
precautions or adverse reactions ... without filing a supplement." That statement
is simply wrong. The section of the regulations he cites deals with supplements,
and the language regarding strengthening warnings, adverse reactions, etc., [21
CFR 314.70(c)] deals with what are termed "changes being effected" (CBE)
supplements that must be submitted no less than thirty days prior to the change
being made. If the FDA does not agree with the proposed changes, it can require
the applicant to change the labeling to its original language. In addition, while
a holder of an approved NDA can submit a CBE supplement for a labeling change as
described, a holder of an approved ANDA cannot unilaterally change its labeling,
since it would make its labeling different from the NDA holder's labeling, which
would be in violation of the FDCA [21 U.S.C. 355(j)], and the FDA would not permit
it.
25. There are several statements made by plaintiffs' expert Michael Hamrell,
Ph.D., with which I disagree and upon which I would like to comment.
* In section II.A. of his report, pp. 6-7, he cites three limitations to the FDA's
ability to monitor the safety of marketed drugs. From my experience at the FDA, I
believe he overstates his case. First he says that the FDA's work is "limited in
that they only have access to data and reports submitted by drug companies to
perform their oversight functions." While this is technically true, it is
important to realize that the law requires that drug companies supply the FDA with
all information about their drugs, including both positive and negative results.
In addition, drug companies and regulatory processes are international in scope.
The FDA is in frequent contact with drug regulatory agencies around the world, as
well as the World Health Organization (WHO). These organizations also have drug
surveillance systems and share data with the FDA. The strength of spontaneous AER
systems lies in the numbers of reports on the wide range of drugs that can be
analyzed to detect signals of problems. The FDA's system, augmented by information
from other systems, provides a resource that no single drug company has. Dr.
Hamrell goes on to discuss the limited resources that the FDA has for analyzing
adverse events. He cites the fact that when he was at the FDA, the drug safety
staff numbered only 52 people. That unit is much larger now, but on pages 6 and
10, he states that reports on drug safety, including periodic reports required of
NDA holders are reviewed only by the drug safety organization, not new drug
reviewers. Those statements are in error. Periodic reports are required to be sent
in duplicate to the FDA's central document room, where one copy is sent to the NDA
reviewing division and one to the drug safety staff. The reviewing medical officer
sees the report as well as someone from the drug safety unit. As I recall, there
are over 100 staff in the drug safety area and hundreds of medical reviewers in
the reviewing divisions. He then makes an allegation that pharmaceutical lobbyists
impede the FDA's ability to do its job of drug safety surveillance. From my
experience, the most important political force exerted on the FDA is the threat of
a congressional hearing if a bad drug gets approved or if the FDA does not take
sufficient steps to correct problems with marketed drugs. For every political
pressure felt at the FDA to approve products there is more pressure felt not to
make mistakes and approve a bad product. Finally, he states that "regulations
provide a minimum set of standards to meet in order to comply. Responsible and
ethical companies usually provide information that not only meets the
requirements, but exceeds them in order to assure that FDA has all the information
it needs to monitor the safety and effectiveness of a drug product." He cites no
support for this statement. However, from my experience, the FDA sets what it
regards as standards needed to produce desired results. The FDA does not regard
its standards, including labeling, as "minimal." Regulated companies must comply
with FDA requirements, and those that do not get into trouble. Drug labeling walks
a tightrope between being informative about potential adverse events (i.e.,
discouraging unsafe use of the product) and being alarmist (i.e., discouraging
safe and beneficial use of the product). The FDA has a wealth of experience with
drug labeling, and it imposes its expertise to make labeling strike an appropriate
balance. More warnings or more dire language in labeling is not necessarily a good
thing, since excessive warnings may deter appropriate patients from getting the
benefits of a drug that could help them. In developing labeling, the FDA seeks to
optimize the balance. It does not permit companies to unilaterally alter that
balance.
* In paragraph 22, page 9, paragraph 28 on page 12 and elsewhere, he repeats a
misconception I have seen before: that FDA regulation 21 CFR 314.80(b) requires
both NDA and ANDA holders to review the world literature on their drug products
and to analyze the clinical significance of AERs. As I discussed in paragraph 12
above, regulation 314.80(b) applies only to NDAs.
* In paragraph 36, page 15, he cites 21 CFR 201.128 for the proposition that drug
companies are responsible for providing adequate labeling for known use outside
the approved labeling of the product. He does not, however, suggest the kind of
labeling language that he thinks should be added. This "intended uses" provision
has been on the books for decades. It is written in such a broad way that, so far
as I can remember, it has never been enforced. I recall a case in the early 1970's
where a major drug company knew that some 90% of its sales of a drug product were
going to an unapproved use for which it was not labeled. Even in that extreme
case, the FDA elected not to take action because the provision was thought to be
unenforceable by the agency's attorneys. It has been difficult for the FDA to
begin enforcing this provision, since, as scientific communication has progressed,
it seems that the ability of medicine to find new uses for drugs has outpaced the
regulatory process for updating drug labeling. In some fields, such as oncology,
off-label use comprises the majority of use of many drugs. In the case of
metoclopramide, however, the labeling does recommend that the product not be used
for more than 12 weeks and does warn that TD is a possible adverse event.
