Plaintiff’s Expert Report in Metoclopramide Lawsuit Explains History of Drug and its Association With Tardive Dyskinesia

This lengthy report explains the regulatory and clinical history of Reglan and Metoclopramide, including their association with movement disorders:

My testimony will include discussion regarding enumerable published reports, that have been accumulating since 1978, that evidence that CAPs are strongly associated with Tardive Dyskinesia (hereafter TD) when given at sufficient dose for a sufficiently long duration (i.e., duration of exposure is a major risk factor for CAP-induced TD). The term “Tardive Dyskinesia” is used to describe all persistent, occasionally reversible, abnormal movements caused by prolonged exposure to dopamine antagonists. The most common form of TD presents as orofacial stereotypy (i.e., involuntary, repetitive oral facial and lingual movements that resemble chewing, lip smacking, tongue protrusion), at times associated with choreatic movements (i.e., brief, irregular contractions that are not repetitive or rhythmic, but appear to flow from one muscle to the next) of the extremities and trunk (Jimenez-Jimenez, 1997).

Apparently, movement disorders have been known side effects of Reglan and Metoclopramide for over thirty years.  Scary.

Technorati Tags: Reglan Lawsuit,Metoclopramide,Metoclopramide lawsuit,Reglan

RE: Martha Pustejovsky v. Wyeth, et al Northern District of Texas, Fort Worth Division Case No.: 4:07-CV-103-Y

Dear Mr. McGlynn:

As you know, my name is Robert C. Nelson, Ph.D., FISPE of RCN Associates, Inc., 160 Main Street, Suite 1740, Stouchsburg, PA 19567, cellular phone 443-454-6484. Thank you for requesting my opinions in Martha Pustejovksy's case. It is my understanding that Ms. Pustejovsky was prescribed and ingested the prescription drug, metoclopramide. It is further my understanding from the prescription records, that Ms. Pustejovsky was given the generic metoclopramide drug manufactured by Pliva, who is a defendant in this action. It is also my understanding that Ms. Pustejovsky has been diagnosed with tardive dyskinesia by her physician. I recently viewed a video tape of Ms. Pustejovsky and in my opinion, agree that Ms. Pustejovsky has tardive dyskinesia. My opinions are contained in my report presented below. The report I submit today is based upon materials that I have reviewed. My opinions are as follows:
I. Background. Qualifications, and Experience
A. Resume and Bioliography

I am a clinical pharmacist by background and initial training. I received my B.S. in pharmacy from Long Island University's Brooklyn College of Pharmacy in 1974, then completed an American Society of Hospital Pharmacist (ASHP) accredited Residency in Hospital and Clinical Pharmacy at the University of Connecticut Health Center. In 1975, I accepted a commission in the U.S. Public Health Service, and served as the clinical pharmacy liaison to the Neurology Institute at the National Institute of Health (NIH).

I transferred to the Food & Drug Administration's (FDA) Division of Neuropharmacological Drug Products in 1977. I continued with the FDA for more than 20 years thereafter. While on active duty, I completed both a Master's in Science (in the Administration of Science & Technology) from George Washington University, and a Ph.D. in epidemiology from the University of Minnesota. I conducted my thesis research, “Psychotropic drugs, psychiatric disease and automobile crashes: a case-control study in 1300 females” at the Mayo Clinic.

At the FDA, I held multiple positions of increasing responsibility and scope in new drug review, epidemiology, and post-marketing surveillance, with an intervening stint to found and direct the Center for Drug's Staff College. In my most recent FDA position as Associate Director (Epidemiology), I lead the re-engineering of the post-marketing surveillance program for human drugs, including utilization of relevant international harmonization initiatives, a comprehensive regulations rewrite, full process re-engineering, and the construction of a new computer system (“AERS”) for spontaneous adverse drug reactions. I retired my commission in the U.S. Public Health Service in February 1998 with the rank of Captain (06).

I am currently President of RCN Associates, Inc., an independent consulting firm in global drug safety, postmarketing surveillance, pharmacoepidemiology, therapeutic risk management, regulatory strategy, drug abuse liability assessments, Good Pharmacovigilance Practices (GPVP), forensic epidemiology, and regulatory affairs. In addition, I advise on the design and functional requirements of software products used in the areas of drug safety, data mining and regulatory document management. I also hold an adjunct Professorship at the School of Pharmacy, University of Maryland. I was elected as a member of the American Society for Clinical Pharmacology & Therapeutics (ASCPT) and also recently elected as part of the initial class of 38 Fellows of the International Society for Pharmacoepidemiology (ISPE).

As a regulator, I was a member of the team responsible for many of the major drug safety issues that arose over the 20 years I served at the FDA. As a result, I developed a deep understanding of the absolute and relative strengths and limitations of all forms of empirical data related to both drug development and drug safety. My major focus was on data from observational research, including adverse reaction case reports.

During my tenure as Director of the FDA/CDER Staff College I was responsible for the training of all professional staff, including all medical reviewers. I oversaw the development of and lectured in courses on regulatory science, statistics, clinical trial design, epidemiology, pharmacokinetics, etc.

My curriculum vitae is attached as Attachment A.
B. Compensation

My opinions and expertise cover the area of general causality and drug regulation. The judgments and opinions in the following report are my own and are made within a reasonable degree of scientific certainty after having reviewed the materials both supplied to me and assembled during the course of my personal research. This report is preliminary. As additional information and/or reports are provided, I will supplement my report as appropriate. For my services, including the review of materials and preparation of this report, I am being compensated at the rate (for clients contracted prior to 2008) of $500 perhour. For depositions and trial testimony I request $750 per hour.
C. Previous Testimony

I have testified via deposition in one metoclopramide case, Kettering v. Wyeth, Cause No. 2003-3118-4, Henrietta Kettering, et al. v. Wyeth, et al., 170th District Court, McLennan County, Texas. I have testified or prepared expert reports in other cases and have listed those cases on attachment B to this report.
II. Purpose of this Review

My opinions and expertise cover the area of general causality, pharmacovigilance, pharmacoepidemiology, drug safety and drug regulation.
III. Materials Used in the Production of this Report

Three general categories of materials were reviewed; those supplied by Lead Counsel, supplemental review of the medical literature, and professional references in my personal library, A schedule of the documents I considered in preparing this report is attached hereto as Attachment C.

The expert report I submit today is based upon the materials that I have assembled and reviewed thus far. I reserve my right to modify these opinions should new materials warrant an update.
IV. Executive Summary

My opinions on this important public health and patient safety matter are as follows: after reviewing the assembled materials presented in this report, it is clear to me that tens of thousands of individuals may be suffering from a situation that is totally preventable. I am stunned that the applicants, both NDA and the ANDA holder (PLIVA), have not rectified the problem over the many years they have been clearly aware of these severe neurologic consequences directly caused by the use, especially the long-term use, of their drug product.

It is my opinion that PLIVA, failed to perform a comprehensive risk analysis of EPS including TD, as well as the NDA Applicant (Wyeth).

It is my opinion that metoclopramide was approved for marketing in Europe (1964) before modern scientific standards for efficacy were established and functional.

It is my opinion that metoclopramide (MCP) was approved for marketing in Europe (1964) before even crude pharmacovigilance systems had been established subsequent to the thalidomide disaster in France and Germany in the early 1960s.

It is my opinion that the evidence supports that MCP is a dopamine antagonist and therefore a neuroleptic.

It is my opinion that the Bradford-Hill criteria are an accepted and tested standard for the assessment of a causal relationship between a pharmacotherapeutic agent and an adverse outcome.

It is my opinion that based on the application of the Bradford-Hill criteria short-term MCP exposure is casually associated with the EPS sub-type known as acute dystonia.

It is my opinion that based on the application of the Bradford-Hill criteria MCP exposure is casually associated with the EPS sub-type known as akathisia.

It is my opinion that based on the application of the Bradford-Hill criteria medium- to-long-term MCP exposure is casually associated with the EPS sub-type known as non-idiopathic Parkinsonism.

It is my opinion that based on the application of the Bradford-Hill criteria long-term MCP exposure is casually associated with the EPS sub-type known as tardive dyskinesia.

It is my opinion that the evidence to support the causal relationship between MCP and TD was sufficient as of 1985.

It is my opinion that the evidence supports that the 1/500 EPS rate was an estimate established for the single-dose inpatient intravenous use of MCP.

It is my opinion that the evidence supports that the Applicant did not revise the EPS section of the product label to reflect the data from the clinical trials performed for the oral dosage form and from the non-U.S. medical literature.

It is my opinion that the 1/500 EPS labeling was inaccurate and misleading from the moment the oral dosage form was approved.

It is my opinion that the evidence supports that acute dystonia is the most frequent (at a rate far higher than 1/500) EPS following acute IV use, especially in infants, children and young adults.

It is my opinion that the evidence supports that Akathisia is the most frequent (at a rate far higher than 1/500) EPS with normal (up to 40mg daily for 8 weeks) dose oral use in adult patients with DG or GERD.

It is my opinion that the evidence supports that older adults exposed to MCP at a normal dose for extended periods are at highest risk for Parkinsonism. Males with decreased kidney function on higher doses are at greatest risk.

It is my opinion that the evidence supports that persons with MCP-induced akathisia are at greater risk for developing TD with continued exposure to MCP.

It is my opinion that the evidence supports that chronic continuous long-term exposure, especially at higher doses, paces the patient at risk for TD.

It is my opinion that the evidence supports that there are high-risk groups for MCP-induced TD, including mentally challenged, females, diabetics, medically fragile, and the elderly.

It is my opinion that the evidence supports that TD (any severity) risk rates for elderly females, diabetic or not, with chronic continuous exposure to MCP may be high and relatively common.

It is my opinion that the evidence supports that if recognized early and mitigated (usually just by discontinuation of offending agent), TD is mild and reversible.

It is my opinion that the evidence supports, to the contrary, that if early TD is not mitigated and exposure continues, the risk for severe, debilitating and irreversible TD is increased.

It is my opinion that the evidence supports that the vast majority (approaching 100%) of MCP is prescribed by non-psychiatrists.

It is my opinion that the evidence supports that non-psychiatrists do not have the APA guided training in the detection and mitigation of TD.

It is my opinion that the evidence and accepted practice supports that estimating incidence when both prevalence and duration is known is consistent with established and tested epidemiological methodology.

It is my opinion that study-level statistically significant results are not required to develop a strong opinion that the prevalence results from the Ganzini, Sewell and Matson research results are valid and reliable for the sub-populations researched, especially since they are consistent with finding from the CAPS and because causally was previously determined using the Bradford-Hill criteria.

It is my opinion that the results from specific sub-populations do not need to be generalizable to the general population to signal a true public health problem.

It is my opinion that the evidence supports that the Food Drug and Cosmetic Act, as amended, and all the implementing regulations have an ethically based patient safety foundation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to monitor and assure that their product is being used safely for intended uses.

It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to monitor the adverse consequences associated with the use of their products.

It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to comprehensively assess signals of risk raised for their products and promptly notify the FDA.

It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to effectively mitigate any identified risk associated with the use, on label or off label, of their products.

It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to assure that their products are accurately labeled for safe use, at all times.

It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that there is no provision of the Hatch-Waxman amendment to the FDCA that overrides the basic safety provisions (1938 and 1962) of that Act.

It is my opinion that the evidence supports that MCP-induced TD due to unapproved off-label use remains an unmitigated public health problem to this day.

I render no opinion on specific causality in this or any other case.
V. An overview of the New Drug Development Process in the United States

The laws and regulations governing pharmaceutical product development are founded on ethical principles and are in place primarily to assure safety. The 1962 amendments to the 1938 Food, Drug and Cosmetic Act (“FD&C Act”) were the efficacy amendments. The modern clinical trial arose only after these 1962 amendments were codified in regulation and implemented. Technically, the FD&C Act prohibits interstate commerce of any regulated product that is not in compliance to the established standards.

Drug development and the FDA approval process involve a complex set of rules, regulation and scientific principles. In order to make it understandable for this report, I chose to truncate the explanation dramatically and to use much of the wording already composed in the series “ From Test Tube To Patient - Protecting America's Health Through Human Drugs” on the FDA consumer website ( http://www.fda.gov/fdac/special/testtubetopatient/html).

A drug company seeking to sell a drug in the United States must first test it. All of these tests are performed by the drug company based on standards issued by the FDA. Before a drug can be tested in people, the drug company or sponsor performs laboratory and animal tests to discover how the drug works and whether it's likely to be safe and work well in humans. The application to test a new drug in humans is referred to as an Investigational New Drug Application, or IND. The FDA will review an IND to decide whether it is reasonably safe for the company to move forward with testing the drug in humans. No drug is absolutely safe; all drugs have side effects. “Safe” in this sense means that the benefits of the drug appear to outweigh the risks.

The testing of a drug in human subjects is called a clinical trial. Clinical trials of a new drug pursuant to an IND are conducted to provide information to the FDA so it can determine whether the drug is safe when used to treat a disease and whether it provides a real health benefit. The trials are conducted in three phases. Phase 1 studies, or clinical pharmacology studies, are usually conducted in healthy volunteers. The goal here is to determine what the drug's most frequent side effects are and, often, how the drug is metabolized and excreted. The number of subjects typically ranges from 20 to 80. Phase 2 studies begin if Phase 1 studies don't reveal unacceptable toxicity. While the emphasis in Phase 1 is on safety, the emphasis in Phase 2 is on effectiveness. This phase aims to obtain preliminary data on whether the drug works in people who have a certain disease or condition. For controlled trials, patients receiving the drug are compared with similar patients receiving a different treatment -- usually an inactive substance (placebo), or a different drug. Safety continues to be evaluated, and short-term side effects are studied. Typically, the number of subjects in Phase 2 studies ranges from a few dozen to about 300.

Phase 3 studies begin if evidence of effectiveness is shown in Phase 2. These studies gather more information about safety and effectiveness, studying different populations and different dosages and using the drug in combination with other drugs. The number of subjects usually ranges from several hundred to about 3,000 people. These are formal hypothesis testing studies.

After obtaining and assembling clinical trial results, the company then sends the FDA's Center for Drug Evaluation and Research (“CDER”) all of the evidence from these tests to demonstrate that the drug is safe and effective for its intended use. A New Drug Application (“NDA”) is the formal step a drug sponsor takes to ask that the FDA consider approving a new drug for marketing in the United States. An NDA includes all animal and human data and analyses of the data, as well as information about how the drug behaves (pharmacokinetics and pharmacodynamics) in the body and how it is manufactured. A team of CDER physicians, statisticians, chemists, pharmacologists, and other scientists review the company's data and proposed labeling included in the NDA. If this independent and unbiased review establishes that a drug's health benefits outweigh its known risks, the drug is approved for sale. It is important to note that the FDA standards are absolute and not relative. This means that (with rare exception) the FDA makes no judgment on the relative merits of one drug when compared to another nor does it make any judgments concerning a new drug's role in the practice of medicine or its viability in the marketplace. The forces that bear on the marketplace are allowed to make those determinations.

If there are problems with an NDA or if more information is necessary to make that determination, the FDA may decide that a drug is “approvable” or “not approvable.” A designation of approvable means that the drug can probably be approved provided that some issues are resolved first. This might involve the sponsor and the FDA coming to a final agreement on what should go on the drug's labeling, for example. It could also involve more difficult issues, such as the adequacy of information on how people respond to various dosages of the drug. A designation of “not approvable” describes deficiencies significant enough that it is not clear that approval can be obtained in the future, at least not without substantial additional data.

NDAs are for specific dosage forms and specified uses (i.e., indications), with a specific product label intended to convey the knowledge obtained in the development of that drug product to the heath professionals who prescribe, dispense and administer the drug. Any changes to that dosage form or label are made through a supplement to that NDA (if, for example, new studies are performed to demonstrate safety and efficacy for a new indication or for a higher dose). If the dosage form itself is modified, that requires the submission of a new NDA. In certain instances that new NDA can reference certain components of an earlier approved application with the approval of the owner of that earlier application.

When the CDER decides that based on all information assembled thus far that is compliant with all recognized standards is sufficient to deem the drug product safe and effective, it issues an approval letter that authorizes marketing. As I have explained in these sections, the NDA process is a continual one and the relative safety and efficacy is constantly re-evaluated in light of new information, usually safety information. Once a drug product leaves the controlled clinical trial environment and enters the clinical practice environment many factors come into play, including the much larger number of people exposed. The importance of the postmarketing efforts and the NDA submissions that are required throughout the marketing life of a drug product is to constantly assure that the benefit outweighs the risk. If new and important information comes to light that calls that balance into question, various risk management tactics and strategies need to be employed, including withdrawal of the application.

Pharmacokinetics (PK) is a branch of pharmacology dedicated to the study of the time course of substances and their relationship with an organism or system. In practice, this discipline is applied mainly to drug substances, though in principle it concerns itself with all manner of compounds residing within an organism or system, such as nutrients, metabolites, endogenous hormones, toxins, etc.

Pharmacokinetics has been broadly divided into two categories of study: absorption and disposition. Disposition is further subdivided into the study of the distribution, metabolism and elimination or excretion of a drug. Thus, pharmacokinetics is sometimes referred to as ADME. Once a drug is administered as a dose, these processes begin simultaneously.

Pharmacokinetics is sometimes contrasted with pharmacodynamics (PD), which is the branch of pharmacology dealing with the reactions between drugs and living systems. Pharmacodynamic effects would include, for instance, any biochemical or physiologic pharmacologic effects of the drug or active metabolites related to the drug's clinical effect in preventing, diagnosing, mitigating, curing, or treating disease, or those related to adverse effects or toxicity. The desired pharmacodynamic effect of a statin, for example, is its ability to favorably alter lipid levels. So, in basic terms, while pharmacodynamics explores what a drug does to the body, pharmacokinetics explores what the body does to the drug.

All clinical investigations of pharmaceutical products, that is, those involving human subjects, are subject to the ethical precepts of the Declaration of Helsinki, ( http://www.wma.net/e/policy/b3.htm) and regulatory Good Clinical Practices (GCP) ( http://www.fda.gov/oc/gcp/requlations.html) and http://www.fda.gov/oc/qcp/preambles/default.htm). In the United States, all regulations that implement the federal FD&C Act, as amended, have an ethical cornerstone. Patient safety is paramount.

It is good practice that the drug development plans for a new molecular entity include a safety plan. A safety plan includes a review of the background epidemiology of the treated disease, its etiology and characteristics, the identification of any safety issues that could be of concern, including events known to be pharmacological class effects, and the detailed characterization or specification of all the identified issues. It also includes a review of the medications used in the indicated and comorbid disease states, here conditions such as obesity or diabetes, for drug-drug interaction potential. These are dynamic plans that evolve as new safety knowledge about the investigational drug is uncovered.

All of these elements are in place to minimize risk to patient safety and conform to universally recognized ethical principles and good clinical practices.

As a component of a new molecular entity safety plan, there would be the details of data acquisition, coding and analyses. Exact methods should be employed over and across all trials in a development program. When there are areas of concern, based on class effects etc., special effort should be made to assure that the data of interest are collected, coded and analyzed appropriately.
VI. The Postmarketing Surveillance of Therapeutic Products in the United States

The same regulatory construct of standard setting, research by the regulated, and monitoring of compliance to standards that existed pre-marketing extend to post-marketing. Patient safety remains paramount. The concept of postmarketing surveillance for safety involves three major components: (i) monitoring labeling and advertising, (ii) monitoring production plants and product quality and (iii) monitoring the use patterns and side effect profiles in clinical practice environments. This report will focus exclusively on the third component.
a. Rationale for postmarketing surveillance of prescription drugs

Randomized clinical (i.e., on humans) trials (RCTs) on pharmaceutical products are experiments designed to maximize internal validity and to test the null hypothesis of no effect. The statistical superiority on one arm (hopefully the test drug product) over the other (usually a placebo) falsifies the null and this finding is considered significant if it is 95% certain, utilizing the appropriate statistical test. The size of a clinical trial is usually determined by statistical power calculations based on anticipated treatment effect. Safety considerations usually do not factor into the determination of sample size. Size of an arm of a Phase III clinical trial can range from a few dozen for oncologic drugs to many thousands for antimicrobial drugs used to treat infectious diseases.

However, the absence of reliable evidence of risk should not be mistaken for reliable evidence of the absence of risk. In the usual case, the clinical trial lacks the power to confidently detect moderately rare events below the order of 1 in 100. The rule of three (Table 8 below) requires 300 persons exposed to provide a 95% certainty of detecting a 1:100 event. Nearly five thousand persons are required in a trial to detect two cases of a reaction that occurs at a 1:1000 rate. Any event occurring more rarely will not be identified within most clinical trials.
b. The Purpose of Postmarketing Surveillance of prescription drugs

The purpose of postmarketing surveillance is to identify, characterize, quantify and communicate both the benefit and the risk, following product marketing, in order to protect the public health with more rational pharmacotherapy. The four key functions of post approval risk assessment are to (1) monitor patterns of medication use, (2) identify and characterize potential new risks following product marketing, (3) assess and quantify known or suspected drug safety issues, and (4) respond to assessed risks. Each function is described below.

i. Monitor the patterns of medication use

The patterns of use can be monitored according to the following categories:

• Each approved indication (e.g., a medicine approved only for clinical depression);

• Off label use (e.g., the same medicine used for seizure disorders);

• Situations of misuse and abuse (e.g., the same medicine used in suicide attempt);

Drug use patterns, while not direct indicators of risk, are essential for the proper interpretation of the context in which the risk occurs. Risks are relative to the circumstances of use. At times, there may be substantially different risk profiles for the different patterns of use. A detected adverse reaction can only be fully characterized within the circumstances of the use, including risk factors of the user and of the drug product. Errors in prescribing, administration, and compliance also affect the overall experience with the safety of a medicine, and should be examined.

ii. Identify and characterize potential new risks following product marketing

This function is fulfilled through the following steps:

Identification of new potential associations between medicines and adverse clinical events, or “signals.” These events include those that may be common, infrequent, or rare.

Assessment of Causation. Assessment is usually conducted through careful evaluation of single cases (such as those reported through the spontaneous adverse experience reporting systems), or case series (including multiple cases reported spontaneously, published in the medical literature, or observed in the clinical setting).

Signal Confirmation. This step involves the assessment of the strength of the signal. It may also involve replication from independent sources, and tests in other models, such as through the techniques of clinical pharmacology.

The ability to rapidly detect, identify and assess new and unanticipated drug safety issues is the cornerstone of the postmarketing surveillance efforts. Progress in spontaneous reporting (MedWatch, FDA's Adverse Event Reporting System (“AERS”), ICH standards, and focused new regulations) is improving our capability in the U.S. to identify signals.

Data systems are available for exploration of signals such as record-linkage data from health maintenance organizations (“HMOs”). These data systems have offered a significant advance in the development of safety assessment tools and have often been used to assess early signals of potential new medication risks. These data are usually collected for administrative or billing purposes and are not dedicated drug safety data, and may therefore not be appropriately detailed to assess all medication-event relationships. Moreover, they are not always amenable to evaluation in the time period required to assess quickly associations of potentially major public health impact

Replication is a scientific standard in experimental research. In observational research, because it has lower internal validity than experimental research, replication is even more important so multiple dedicated data sources are required for evidence formation.
iii Assess and quantify known or suspected drug safety issues

Understanding the suspected risk and placing it into appropriate clinical context is perhaps the most challenging aspect of postmarketing safety assessment. This process requires the following elements:

Description of the risk. The frequency, or incidence, of the adverse event is measured according to the relevant use of the medicine (e.g., number of events per 1,000 persons taking the new antidepressant for clinical depression). The frequency is evaluated according to factors that may influence the occurrence of the event. These risk factors include those associated with demographic or metabolic characteristics of the patients, as well as the concomitant administration of drugs, foods, and alcohol.

Quantification of the risk. The frequency of the risk is placed into context by calculating the relative risk (e.g., how often does the event occur with this medication relative to other medications for the same medical condition, or relative to no therapy). The frequency may also be calculated as an absolute risk (e.g., the proportion of the total number of people taking this medicine experience this event).

Data collected systematically during clinical trials provides a strong basis for the initial use of a newly marketed drug product in clinical practice. However, as clinical practice exposure broadens there are no data systematically collected which will monitor the continued accuracy (or inaccuracy) of those clinical trial estimates. Data that can both describe and quantify the risk profile of a drug are required. Drugs that are grouped within a class, e.g. the statins, that are in reality very different compounds with different pharmacological activity need to be assessed as individual agents in order to determine if the rates of known ‘class-effects' are different to a meaningful extent. Some computer databases exist that are both large enough and detailed enough to quantify risk and evaluate risk factors, such as those of HMOs. Often these databases must be supplemented with additional data collection to validate or supplement computer information with detailed information from patient medical records.
iv Respond to Assessed Risk

Once a new risk has been identified, evaluated and placed into context, that new information is communicated to the following groups:

• Regulators

• Healthcare practitioners

• Consumers and the media

In addition to communicating the risk, additional action is sometimes required. That action may include re-labeling the medicine (i.e., changing the package insert), restricting the product's use or indications, or withdrawing the product from the marketplace. Regulatory mechanisms currently exist which can limit continued exposure to the risks identified. A variety of labeling changes and withdrawals from the marketplace are available. In addition, the medical profession can, through peer reviewed and other publications, educate practitioners on the merits of various pharmaceuticals. These methods are currently hampered or delayed due to weak or conflicting safety data. Moreover, the U.S. label also has at least two major shortcomings: 1) most changes are not highlighted from version to version, and 2) there are weak criteria for how most adverse drug reactions are listed.
c. The Post Marketing Surveillance Toolbox

The required functions of postmarketing drug risk assessment outlined above require differing capabilities to address each of the differing requirements and questions that arise. No one tool can be used to address all the issues. The identification and characterization of rare and unexpected new risks are the purview of spontaneous reports. Spontaneous reports can be a powerful tool for this purpose.

Various types of reaction detection schemes are required to address the types of data that are necessary to assess the safety of a drug in the marketplace comprehensively. These data sources, and thus the capability to perform certain risk assessment functions, vary between the U.S. and the European Union (“EU”). An easy demonstration of this is that in the United Kingdom (“UK”), in addition to the VSRS (their the Yellow card scheme; enabled by their ADROIT computer system), they developed a Green card scheme which follows up on populations of new drug users (run by the Drug Safety Research Unit, Southampton, UK) and the General Practitioners Research Database (GPRD) which monitors the prescribing and adverse outcome in a sample of the UK's general practitioners. The FDA has only the VSRS (enabled by their AERS system) in place. It has failed for 40 years to construct other tools, including rigorous enforcement of PMS regulations, and properly protect the public health. The FDA thus relies on the applicant, which bears the responsibility to develop the tools that are required to assess safety signals that arise.

In addition, programs run by academic or for-profit organizations exist. Most involve the use of secondary data sources (i.e., administrative or billing records, medical records, pharmacy records) for epidemiological research. The strengths and very substantial limitations of these data are well documented. A summary of these capabilities appear in the three major textbooks in this field (Hartzema et.al., 1991; Stephens et.al., 1998; Strom, 2000). Most of the funding for these resources comes from the pharmaceutical industry. A model for regulatory decision-making using a variety of observational data sources was developed and a number of case examples are illustrated in Nelson 1998.

The FDA has recognized these deficiencies and has performed a recent capability assessment in 1999 and developed a conceptual framework for the management of risks. The message is to develop educational and risk communication materials to aid the prescriber and patient to prevent to avoidable outcomes. In addition, they recognized the need to quantitate the unavoidable, and to identify then quantitate the currently uncertain outcomes. Overall, this effort is referred to as the FDA's Risk Management initiative.

Most recently, The Prescription Drug Users Fee Act --IV (PDUFA-4), has increased staffing and funding for the FDA's postmarketing surveillance program and developed good risk management procedures for the industry. However, until these new initiatives produce the required types of dedicated drug safety data, risk assessments on pharmaceutical products will continue to be made on those secondary data that are available.
d. Current US Regulatory Framework

New drug applicant holders, both NDA and ANDA, are required to collect, assess and submit ALL new safety data at a variety of levels. They generally include: the individual spontaneous report, the case-series, literature of pharmacoepidemiological research results for labeling changes, and the periodic report. The individual case is submitted either expedited (submit within 15 days) if it fulfills those requirements under 21 CFR 314.80 or periodically. Evidence supporting labeling changes are to be submitted whenever new information warrants. The applicant is ethically bound to actively acquire the required information to characterize the safety profile of their drugs in the clinical practice environments. Then and only then is maximal patient safety and public health possible. The periodicity of the NDA postmarketing safety update is based on the time since approval. The contents and format of the periodic submissions will be changed when the FDA's recently proposed rule becomes final and facilitates international harmonization, but the ethical underpinning remains the same. Product labeling needs to be reflective of all known safety issues to avoid the drug being considered misbranded under the FD&C Act.

The general scope of the applicant's responsibilities is stated in 21 CFR 314.80 (b): “Each applicant having an approved application under Section 314.50 or Section 314.55 shall promptly review ALL drug experience information obtained or otherwise received by the applicant from ANY source, foreign or domestic, including information derived from commercial marketing experience, postmarketing clinical investigations, postmarketing epidemiological/surveillance studies, reports in the scientific literature, and unpublished scientific papers.” A more extensive overview of the U.S. regulatory framework appears in Nelson 1998, and Kennedy 2000.
e. Industry ADR Process

As they did prior to approval, all new drug Applicants, holders NDA or ANDA, have the continued responsibility for assembling data and assessing the safety of their products post approval, in various relevant clinical practice environments. Some of the tools used in this “pharmacovigilance” activity are the published literature, results of ongoing clinical trials, dedicated drug safety data collection schemes, focused clinical pharmacology investigations, observational research, and spontaneous adverse reaction reports. All information relative to the “benefit -- risk balance” must be assessed by the applicant and sent to the FDA. Maximal reporting time periods are set by regulation.

Most application holders establish drug safety units within their organization that have automated databases for spontaneous safety reports, and follow a set of regulation-compliant standard operating procedures (“SOPs”). Proper documentation is essential, especially of the judgments made on safety issues.

The corporate drug safety department is usually responsible for the receipt and handling of spontaneous reports, their entry into a database, and their submission to the worldwide regulators. They also maintain staff to review, assess and search for signals within those incoming data. A signal is broadly defined as any collection of new data that requires in-depth assessment. Those ‘worked-up’ signals, if demonstrable of new or substantially suspected information, need to be conveyed to the regulators with comments on appropriate risk management tactics, such as a labeling change and/or Dear Doctor letter. Timely and thorough assessments submitted to the regulators allow a company to discharge its duty to warn and protect the public health. The corporate regulatory affairs and legal departments often are involved when labeling submissions are warranted.

Therefore, an applicant is bound by ethics and regulation to actively acquire, assess and report all evidence, at the information level, which could impact the prescribing decision process.
f. FDA's Current ADR process

The responsibility for the review of post-marketing safety issues in the CDER are shared between the risk assessors in the Office of Drug Safety (“ODS”) and the risk managers in the office of Drug Evaluation's reviewing divisions.

The ODS receives and handles all the spontaneous reports from the public directly (about 10-15%) (via the MedWatch program) and indirectly (about 85-90%) from the manufacturers. They place them into a database referred to as the AERS. The AERS system is described in some detail in Nelson, et.al.1998, and Bright et al, 2001. The AERS system alerts the ODS risk assessor when new spontaneous reports have been received that meet certain predefined criteria. They review these reports for ‘signals', then review those signals in the context of basic knowledge, the literature, natural history and epidemiology of disease, and other observational data sources. They also review the submissions (supplements and periodic safety reports) from the company. These NDA submissions contain all materials, from all sources, related to the safety of that marketed drug, including the results of all investigations, the literature, and the spontaneous report summaries and assessments. The content of theses periodic submissions serve as the companies' assessment of the safety of their product, worldwide. A recent overview of FDA drug safety program & processes is found in Kennedy, et.al. 2000

The FDA has historically devoted only a small dedicated staff to post-marketing safety issues, usually no more an 20 - 40 health professionals and scientists for over 20,000 drug products.
VII. The Responsibilities of an Applicant

Listed below are selected sections of the regulations relevant to this matter:

201.128 - Meaning of “intended uses”.

The words intended uses or words of similar import in §§201.5, 201.115, 201.117, 201.119, 201.120, and 201.122 refer to the objective intent of the persons legally responsible for the labeling of drugs. The intent is determined by such persons' expressions or may be shown by the circumstances surrounding the distribution of the article. Labeling claims, advertising matter, or oral or written statements by such persons or their representatives may for example, show this objective intent. It may be shown by the circumstances that the article is, with the knowledge of such persons or their representatives, offered and used for a purpose for which it is neither labeled nor advertised. The intended uses of an article may change after its manufacturer has introduced it into interstate commerce. If, for example, a packer, distributor, or seller intends an article for different uses than those intended by the person from whom he received the drug, such packer, distributor, or seller is required to supply adequate labeling in accordance with the new intended uses. But if a manufacturer knows, or has knowledge of facts that would give him notice, that a drug introduced into interstate commerce by him is to be used for conditions, purposes, or uses other than the ones for which he offers it, he is required to provide adequate labeling for such a drug which accords with such other uses to which the article is to be put.

201.56 9a)(1) - The labeling must contain a summary of the essential scientific information needed for the safe and effective use of the drug.

312.32(b) - IND review of safety data - requires the sponsor to promptly review all information relevant to the safety of a product from any source including ‘reports in the scientific literature and unpublished scientific papers'

314.150 (2) (i)

FDA finds; That clinical or other experience, tests, or other scientific data show that the drug is unsafe for use under the conditions of use upon the basis of which the application or abbreviated application was approved; or

314.150 (b) (10)-(10) That the labeling for the drug product that is the subject of the abbreviated new drug application is no longer consistent with that for the listed drug referred to in the abbreviated new drug application, except for differences approved in the abbreviated new drug application or those differences resulting from:

(i) A patent on the listed drug issued after approval of the abbreviated new drug application; or

Exclusivity accorded to the listed drug after approval of the abbreviated new drug application that do not render the drug product less safe or effective than the listed drug for any remaining, nonprotected condition(s) of use.

Under 21 CFR 314.80 and 21 CFR 314.97 - 314.98, applicant holders are required to collect, assess and submit ALL new safety data at a variety of levels. Generic products or NDA'ed OTC products are covered by the same scope of regulation as the innovator applicant.

FDA regulations require all applicants to provide labeling that contains the essential information necessary for safe and effective use of the drug, and mandates that a prescription drug's labeling must be accurate and not false or misleading. 21 CFR 201.56, “General requirements on content and format of labeling for prescription drugs” states that:

The labeling shall contain a summary of the essential scientific information needed for the safe and effective use of the drug. The labeling shall be informative and accurate and neither promotional in tone nor false or misleading in any particular.

The purpose of this regulation is to protect the public health by ensuring that physicians prescribing drugs will have the information they need to accurately asses the benefits of the drug and to understand the risk of side effects. The regulatory requirements on the content and format of labeling for human prescription drugs apply equally to generic manufacturers as they do to brand name manufacturers, since they are both considered applicants under the FD&C Act, as amended. The regulation, 21 CFR 201.56 imposes an independent duty upon generic manufacturers, this includes PLIVA.

The safety responsibility of an Applicant under 314.80 and 314.81 does not distinguish between NDA and ANDA holders. Both are responsible for assuring that their product is safe and effective in the marketplace. 21 C.F.R. §314.80(b) requires PLIVA to “promptly review all adverse drug experience information obtained or otherwise received by the applicant from any source, foreign or domestic, including information derived from commercial marketing experience, post-marketing clinical investigations, post-marketing epidemiological/surveillance studies, reports in the scientific literature, and unpublished scientific papers”).

This duty applies to PLIVA and the innovator. See 21 C.F.R. §314.98. Based on the materials that I reviewed, PLIVA failed to monitor the safety of its drug. PLIVA, in other MCP litigation, claimed that the FDA was aware of the association between MCP and TD since the mid-1980s, but the risk was assumed to be rare, especially given the preponderance of short-term usage. PLIVA did not take any action to correctly inform the FDA about the actual risk which appeared to be common in certain risk cohorts.

It is my opinion that there is no provision of the Hatch-Waxman amendment to the FDCA that overrides the basic safety provisions (1938 and 1962) of that Act.
VIII. Brief MCP Regulatory and Marketing Background

Development of MCP began in the 1960's and was marketed by Laboratories Delagrange as an acute antiemetic in France in 1964. Note: this was prior to any efficacy requirements for new drugs in France. Merck explored MCP (named MK745 during IND phases) in the U.S. and sponsored the Bodi (1966) study that confirmed substantial EPS with the oral dosage form. Merck subsequently abandoned MCP development.

There were dozens of EPS reports from the European literature in the 1960-70s. It is important to note that the submission or even the acknowledgement of foreign literature was not an FDA requirement until 1985. The Irish Ministry of Health published a warning letter to all physicians regarding dystonia and EPS (having received 33 spontaneous reports) in 1976.

A.H. Robins licensed the U.S. marketing rights to MCP from Delagrange. They also needed to conduct new efficacy trials for any indication for use in the U.S. (Note: from 1962 through the late 1990s, the FDA did not accept foreign studies as grounds for ‘substantial evidence’ for efficacy).

Reglan was first approved in U.S. on February 7, 1979 for:

“Single doses of metoclopramide may be used to facilitate small bowel intubation (e.g., for biopsies) in adults and children in whom the tube does not pass the pylorus with conventional maneuvers.

Single doses of metoclopramide may be used to stimulate gastric emptying and intestinal transit of barium in cases where delayed emptying interferes with radiological examination of the stomach and/or small intestine”

The oral tablet was approved in 1980 for limited use for the treatment of acute diabetic gastroparesis with a controlled efficacy database of 35 persons. It was subsequently was approved in 1984 for the treatment of gastroesophageal reflux disease (GERD) (commonly known as “heartburn”) after a number of FDA rejections, and for only limited use (maximum of 8-12 weeks.)

Although chemically distinct from the conventional anti-psychotics (hereafter CAPs) like chlorpromazine, MCP, like the CAPs, is a dopamine receptor antagonist and therefore a neuroleptic. The American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM IV, pg. 679) reads “The term neuroleptic is used broadly in this manual to refer to medications with dopamine-antagonist properties. These include the so-called “typical” antipsychotic agents (e.g., chlorpromazine, haloperidol, fluphenazine), “atypical” antipsychotic agents (e.g., clozapine), certain dopamine receptor blocking drugs used in the treatment of symptoms such as nausea and gastroparesis (e.g., proclorpromazine, trimethobenzamide, thiethylperazine, and METOCLOPRAMIDE), and amoxapine, which is marketed as an antidepressant”.

Therefore, concern about extrapyramidal adverse reactions (hereafter EPS) existed long before MCPs U.S. development. MCP has been associated with a broad spectrum of EPS. However, if MCP was used only for the brief times indicated, the most severe sub-types of EPS, such as tardive dyskinesia, could be almost completely avoided. Upon the expiration of the patent exclusivity, several drug companies, including PLIVA, began manufacturing generic versions of MCP and soon dominated the market.
IX. Fundamental Methodologies for Assessing Causation from Multiple Information Sources

General causation is the scientific determination as to whether a pharmaceutical is capable of causing a specific adverse outcome in some portion of the exposed population at some dose or conditions.

Many notable adverse drug outcomes that develop during clinical practice are not discoverable in the new drug approval process because they result from drug-drug interactions, present under non-examined circumstances, or are sufficiently rare. Therefore, in the postmarketing assessment of general causality, the randomized clinical trial is an infrequently available tool. In addition, clinical trials often focus on a single factor, when in clinical practice environments most disease processes are multi-factorial. Lastly, they can present their own ethical and moral limitations. Therefore, less internally valid forms of research and data are used to assess these low, rare, but often of public health importance, risks.

Sir Austin Bradford-Hill (1965) set forth nine criteria that have been recognized by the scientific community at large as the method for assessing and establishing general causation. These criteria need to be considered before deciding the most likely interpretation of its causation. These criteria have been widely reviewed in the peer-reviewed literature (Evans 1976; Hackney & Lynn 1979; Doll 1984; Strom 1994; Shakir & Layton 2002; Talbot & Waller 2004). They have also been used by scientific bodies such as the World Health Organization (WHO) the US Public Health Service, the US Environmental Protection Agency (EPA) and the National Academy of Sciences. The most famous early work using these criteria was the 1964 US Surgeon General report on smoking and lung cancer (US DHEW 1964). I concur with the consensus position held by the scientific and medical community that the Bradford-Hill criteria should be used as the cornerstone of the scientific method for determining general causation (Guzelian et.al. 2005).

The recent CIOMS IV report (1998) also offers a framework for the assessment and balance of benefit --to --risk of marketed pharmaceuticals. Their work incorporates the Bradford-Hill criteria.

The Bradford-Hill criteria are: strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experimental evidence and analogy. None if these is absolutely necessary for an association to be causal. Also, no one is sufficient for an association to be considered a causal association.

In assessing a drug-outcome relationship it is necessary to examine the multiple types of data and information resulting from basic pharmacology, animal studies (aka preclinical toxicology), clinical pharmacology studies (including all Phase I clinical trials), safety and efficacy from experimental clinical trials (including all Phase II and III studies), safety from observational clinical trials (or pharmaco - epidemiological research), postmarketing spontaneous reports, and information on drugs of the same pharmacological class and the same therapeutic class.

These information sources need to be placed in context of the nine Bradford-Hill criteria, which are:

Strength: This first criterion states that strong associations are more likely to be causal than weak (i.e., less than RR = 2) associations. That is because weak associations are more likely to be explained by unrelated biases that are factors in observational research. This is a major issue in pharmacoepidemiology since most large risks are discovered prior to marketing so therefore most new risks are weak and hard to detect.

Consistency: Analogous to replication in experimental research, in observational research repeated observations of an association in different populations under different circumstances provide support for a causal association. Because of the low levels of risk generally detected in pharmacovigilance and pharmacoepidemiology, the consistency of findings in different information types is highly important. However, at times specific drug-outcome events only occur under defined conditions. The utilization of this consistency criterion is the most robust approach to assessing any uncertainties resulting from the low risks that commonly occur in the field of drug safety.

Specificity: Bradford-Hill stated that a cause leads to a single event, not multiple effects. This also refers to the question of whether the cause ever occurs without the presumed effect and whether the effect ever occurs without the presumed case (Strom 1994). This criterion is rarely met but when it is strong evidence of causality is provided.

Temporality: Simply stated this criterion requires that exposure (i.e., the cause) precedes the effect in time. It also states that a consistent temporality is stronger evidence than an inconsistency. So in general, an event that consistently occurs 4-6 days after exposure is more supportive then one that occurs anywhere from 4-30 days later. The pharmacological characteristics of the drug and host response are factors that are considered within this criterion.

Biological gradient: This criterion is more commonly termed the ‘dose-response curve’. This states that the size of a dose and/or its cumulative dose and/or the length of exposure correlate with increasing risk (e.g. Smoking and lung cancer). In pharmacovigilance, a causal association is supported when an ADR occurs in a dose-dependant (hereafter, dose-response relationship) manner or from cumulative exposure (hereafter, duration-response relationship) over a prolonged period of time.

Plausibility: Bradford-Hill stated that biological plausibility is supportive but the lack of plausibility cannot rule out the association because we may yet be ignorant of the mechanism of action (e.g. ACEI and cough) or they may be Type B ADRs, or the outcomes have a much-delayed onset.

Coherence: Coherence of a cause-and-effect interpretation or causality assessment should not have data in serious conflict with generally known facts of the natural history and biology of the disease. In other words, does it make contextual sense? The same caveat exists as for plausibility, that is, the current understanding may change as science progresses.

Experimental evidence: In pharmacovigilance this usually refers to returning to the experiment, such as a clinical pharmacology study, to further explore a signal or an association.

Analogy: In pharmacovigilance, the criterion of analogy refers to class effect, be it pharmacological or therapeutic.
X. General Causation for Metoclopramide and Tardive Dyskinesia

My testimony will include discussion regarding enumerable published reports, that have been accumulating since 1978, that evidence that CAPs are strongly associated with Tardive Dyskinesia (hereafter TD) when given at sufficient dose for a sufficiently long duration (i.e., duration of exposure is a major risk factor for CAP-induced TD). The term “Tardive Dyskinesia” is used to describe all persistent, occasionally reversible, abnormal movements caused by prolonged exposure to dopamine antagonists. The most common form of TD presents as orofacial stereotypy (i.e., involuntary, repetitive oral facial and lingual movements that resemble chewing, lip smacking, tongue protrusion), at times associated with choreatic movements (i.e., brief, irregular contractions that are not repetitive or rhythmic, but appear to flow from one muscle to the next) of the extremities and trunk (Jimenez-Jimenez, 1997).

Metoclopramide (MCP) is a dopamine receptor-blocking drug (DRBD), or a neuroleptic. MCP blocks D1 and D2 receptors centrally in the chemoreceptor zone (antiemetic) and peripherally in myenteric neurons (prokinetic). MCPs' other pharmacological properties include 5-HT3 receptor antagonism, 5-HT4 receptor potentiation, and sensitization of muscarinic receptors (Pasricha 2006).

MCP is a neuroleptic (DSM IV, pg.467) that has no FDA approved indication as an antipsychotic and is not usually used as an antipsychotic. However, MCP does have antipsychotic properties at much higher doses then generally employed for the labeled indications. Confirming this, Neuroleptic Malignant Syndrome (NMS) has been associated with the use of MCP (Nomino 1999; Nachreiner, 2006)

Almost all inpatient (i.e., hospital) use of MCP is as an anti-emetic, and is administered acutely, usually a single or a few doses. Other inpatient uses include its original indication, and the more serious cases of DG. Most (~85-90%) of outpatient use of MCP is acute, short term or intermittent (Schaffer 2004). Almost all inpatient and outpatient prescribers of MCP are non-psychiatrists. The EPS subtype-profile varies with dose and duration of use with acute dystonias being most frequent in acute exposure states.

The first evidence of EPS with MCP appears to have been found in the one and only controlled long-term safety study, that I am aware of, conducted with MCP. The Bodi (1966) long-term safety study on 42 prisoners presented 3 episodes of Parkinsonian symptoms and 2 episodes of dyskinetic movements for a 12% incidence rate over three months or 1 in 8 EPS incidence rate. This differs markedly from the 1/500 rate subsequently listed in the product label. Then sponsor, Merck, soon abandoned MCP development, likely because of these findings. The depth of ADR detection and data collection in Bodi stands in contrast to the casual and passive efforts in the multi-centered trials that comprised the modest safety database in the NDA for the oral tablet.

Tardive dyskinesia is an involuntary movement disorder that is always drug induced and was first described in the late 1950's following the introduction of the major tranquilizers, now called CAPS. The term tardive dyskinesia was first used in 1964. Quoting Orme (1984), “the tardive dyskinesias are a group of neurological disorders producing persistent, repetitive, choreiform movements of voluntary muscles, most frequently of the faciolingual-buccal muscle groups, where they resemble sucking, chewing, or lip-smacking. They tend to disappear during sleep and are made worse by emotional stress and voluntary movements of the limbs. The symptoms may be mild and more upsetting to observers than to patients, but they are often severe, distressing, and incapacitating. Though by definition tardive dyskinesia occurs after medication with neuroleptic drugs, it often present after treatment has been stopped”.

Case reports in the literature and evidence of causality for the MCP-TD relationship from case-series reports that had been accumulating since 1978, a year before MCP was marketed in the U.S.

Lavy (1978) reported the first case of TD. A 48 year-old male on high doses for 6 years suffered TD when he discontinued MCP. Symptoms improved when MCP was restarted, that is, a positive rechallenge. Symptoms emerged again at a second attempt to withdraw MCP and subsided when MCP was rechallenged again.

That case was soon followed by 3 cases by Kataria (1978) in the United Kingdom, where MCP was used widely for nausea, vomiting and “gastrointestinal symptoms”. The first case was a 76 year-old female on 30mg/day MCP for 6 years. At, 51/2 years she developed signs of TD, which progressed in severity as she was continued on MCP. Three weeks following discontinuation, most symptoms resolved. The second case involved an 86 year-old female on 40mg/day MCP who developed TD after one year of continuous exposure. She also developed PD. Upon discontinuation, most signs decreased by 3 months but the copulatory dyskinesia (i.e., pelvic thrusting) persisted for 9 months. The last case was also female (age = 66) on only 15mg/day who developed TD after two years, then signs of PD and worsening TD but remained on drug for a total of 4 years. Her symptoms improved after 9 months drug-free. The authors concluded that MCP causes chronic TD and it should join the list of dopamine antagonists known to do so. They further offered that long-term use should be avoided, if possible.

The Swedish Committee for Drug Side Effects (Haggstrom, 1981) warned of the risk of TD in the elderly treated with MCP. They reported that in addition to over 30 cases of MCP-induced acute dystonia and other EPS that have occurred mostly in the young, they now have 5 cases of TD in the older treated patients. Details on three were given (all females, aged 71,83 & 76; duration = 4, 9 months and several years). These MCP-TD cases raised a safely signal as they continued to accumulate. They served as the basis for the 11 cases analyzed by Wiholm (1984)

Grimes and his Canadian colleagues made three early reports on this public health issue. In 1981 they reported in the NEJM that while MCP was recently approved in the U.S. for limited single-dose IV indications, oral MCP was marketed in Canada since 1974. He witnessed 12 cases of MCP-induced Parkinsonsim and seven cases of TD. The TD patients were on long-term therapy with a mean of 2.5 years. He was in the process of evaluating three more TD cases. He warned that both DG and GERD indications could lead to long-term use of MCP and it's potentially irreversible neurological consequences. In a follow-up letter in Lancet, Grimes (1982a) they report that their tally is now 12 cases of TD with a mean age of 72, mean dose of 29mg and mean duration of 26 months (range 8 -60 months). Eight of these patients had persistent symptoms for 15 months following discontinuation of MCP. They advise that prolonged use in the elderly should be avoided. This article was published prior to the U.S. approval fro GERD! They describe their experience with MCP in more detail in the Canadian Medical Journal (Grimes, 1982b).

Indo and Matsuoka (1984) report TD in a 64 year-old male on 30mg MCP for 9 months for a Gl disorder. They postulate an age differentiation on MCP-induced EPS sub-types. Orme (1984), was one of the first to describe the epidemiology of the EPS sub-types induced by MCP. He emphasized prevention since, once TD is established, there is no effective treatment. He advised against long term use and use for trivial symptoms.

Wiholm (1984) was the first to explore the signaled problem. His results need to be understood in context. When oral MCP was launched it was not expected to be associated with TD, because of its' labeled dose and short term indications. However, spontaneous case reports and literature reports for MCP-TD appeared as early as 1978. The signal was raised (Haggstrom 1981) for MCP-TD in the Swedish National ADR database indicating that TD may occur and it may occur less than rarely. Wiholm explored this same database to describe the case-series for risk factors and to broadly estimate frequency of occurrence. In his published report, he described 11 cases, most (7 beyond 6 months) of which clustered upon long-term use and in elderly females. He found no cases amongst short-term users, as expected. Therefore, there was clearly an increased risk with long-term exposure analogous to that found with the CAPs. He estimated both the frequency of ‘off-label’ LT use and attempted to get an upper bound estimate of incidence. He found that 11% (2/3 of which were older than 70) were long-term users and that incidence, if it could be calculated, would be greater than (i.e., occurring more frequently than) 1:1000 in the elderly which was far greater than near-zero risk purported amongst the nearly 90% of persons exposed to short-term therapy.

It is my opinion that the Bradford-Hill criterion for a casual relationship between metoclopramide and tardive dyskinesia was established by 1984. The quantification of the risk is not a necessary criterion for the determination of risk.

All the above case report data should have been offered to and considered by the U.S. FDA prior to the approval of the GERD indication in 1984. At this point, it is my opinion that had a comprehensive risk assessment been performed by the Applicant they would have seen that there was more than enough evidence to modify the product label to independently add a section on TD and it's risk factors. In addition, risk mitigation steps could have been taken to prevent the use of MCP in clinical situations where there was no evidence of benefit and evidence of escalating risks. One of those risk mitigation steps may have been to NOT approve MCP for GERD.

The risk of TD was added to the MCP label in 1985/86 (along with the class labeling for CAPs) by FDA request, without a thorough FDA review or a comprehensive assessment of LT-MCP-TD by the Applicant. Thus, when PLIVA began marketing MCP, it was aware of both LT use (via IMS data and Market survey) and 26 U.S. cases as an adverse consequence of that exposure (Board 1986 NEJM). Published empirical evidence (Wiholm, Stewart) of unapproved LTT (up to 1/3 in specific sub-populations) with MCP has been in the literature since, at least, 1992.

The Brietbart (1986) letter in the NEJM reported TD following high-dose IV MCP (six seven day courses following cancer chemotherapy) in a 45 year-old male with no history of a movement disorder or another neuroleptic. The TD manifested 3 days after the sixth course and persisted for 3 months until the patients' death. He cites other literature on high-dose MCP that provided akathisia and acute dystonia rates of 33% in children and 27% in young adults, versus 2% in older adults. Board (1986) of A.H. Robins responded with a responding letter citing the 26 cases of MCP-TD in the corporate database (20 were suitable for evaluation) stating that most occurred AFTER three months (mean = 13 months) on therapy, including three cases with onset soon following discontinuation. The average age in these reports was 67 years with 15/20 being female, and on normal doses. Dr. Board's correspondence made it clear that the Applicant was aware of the problem of both long-term therapy and it's TD consequences.

A case of severe TD in an elderly female due to a 2-year chronic exposure to MCP was reported by Beauclair & Fontaine (1986). Sarnie (1987) reported a case of life-threatening TD caused by MCP in a 66 year-old male. The TD was misdiagnosed for many months by a number of different physicians and MCP was continued. Despite discontinuing MCP the TD was so severe that a gastrostomy tube was required for nutrition. Also in 1986, Lazzara et.al., reported another elderly female treated with MCP for 8 months at 30mg/day and offered guidelines for the safer use of MCP in the elderly.

They suggested that MCP be used only as indicated, at the lowest effective dosage and to consider brief periods off medication. They further suggested that physicians need to monitor for the earliest signs of EPS and consider discontinuation of MCP.

Lang (1989) described the clinical differences between MCP and CAP induced TD. While most symptoms were shared, the occurrence of pronounced pelvic thrusting and respiratory dyskinesias were substantially more common in the MCP treated patients. Sewell et.al., (1992a) present two case reports in hemodialysis patients with DM who developed MCP-induced TD within 12 months of beginning treatment for DG. They discuss that non-psychiatrists ware less likely to detect and diagnose early TD.

Sewell et.al. (1992b) went on to review and analyze 67 cases of MCP associated TD from the published (N=21) literature. They found that the average time to onset was 20 months and that in 15/21 patients followed long term, it persisted 6 months or more. Interestingly, 20% of these patients took LT MCP for nausea and vomiting. They considered persistent MCP-induced TD both serious and largely preventable. Relevant portions of Sewell's table is reproduced below:
TABLE

Note: These were in addition to the 26 cases that the Applicant had in their database (Board, 1986)

There is no scientific evidence for MCP's effectiveness for any indication with long-term therapy (Lata and Pigarelli 2003). Proton-pump inhibitors (hereafter, PPI) and other therapies are considered by GI experts (e.g., Freston deposition) as the drugs-of-choice for GERD. Given that, there is no established benefit to the long-term use of MCP. MCP is the only drug currently approved for short-term gastroparesis. If an individual physician decides to use MCP to treat chronic (diabetic or non-diabetic) gastroparesis under the practice of medicine, he/she maybe able to minimize the TD by careful neurological symptom monitoring and dose/duration adjustments. However, prior to applying this mode of practice he/she must understand that the prevalence of TD with LT-MCP pharmacotherapy could be 13-40% in certain sub-populations, as well as understanding how to recognize and minimize (e.g. APA Guidelines) the nature and severity of TD. To do so under the current labeling presents an unrecognized (by the practitioner) danger.

MCP-induced TD has also been reported in infants and children. Putnam (1992) reported about an 8 year-old boy treated 7.5 months for GERD and having symptoms during treatment and for 15 months after discontinuation. Mejia and Jankovic (2004) describe TD in a 1 year old infant treated with a supratherapeutic dose of MCP for 17 days for gastroesophageal reflux. The intensity increased for a month after discontinuation.

There is no substantial evidence that MCP is effective for treatment of any disease state beyond short-term application.

The evidence is very clear and strong that MCP is a neuroleptic that induces TD when a patient is exposed to a sufficient dose for a sufficient duration.
XI. MCP induced Extrapyramidal Reactions

In addition to the early (Bodi, 1966) recognition that MCP induced a spectrum of EPS, were many published studies and reports that there were differing types of EPS occurring under differing circumstances to differing sub-populations of exposed persons. The first published report (Witzel, 1968) on MCP-induced acute dystonic EPS was 1968, in children.

The safety data from the NDA for the acute DG are presented below:
TABLE

The entire controlled study database consisted of 35 subjects in which 4 (11.4%) presented with complaints that were likely akathisia. The open and uncontrolled studies there was only one complaint of EPS but it was severe enough for the patient to dropout. Note that the quality of safety data ascertainment and collection are suspect when one (Study #3) reported no ADRs of any kind. The EPS rate of 11.4% was far greater than 1/500 that remained in the labeling. To note, gastroenterologists performed all these studies and they were reviewed at the FDA by a gastroenterologist (T. Garvey) at a time when the general awareness of akathisia was low. The indication of acute DG was approved with this small amount of evidence because it was intended to be limited to only those patients with disease severe enough to require hospitalization. It needs to be noted that this indication was approved with the contingency that the drug be limited to hospital pharmacies and that the Applicant monitor that it use remains compliant to the limited indication. In the material that I have reviewed I have not seen any subsequent efficacy or safety data that supported the broadened use in outpatient and general practice.

One of the clinical trials used for the NDA approval (GERD indication) was published in 1977 (McCallum et.al.). Using a 10mg QID daily dose they reported one patient with marked nervousness and anxiety (i.e., possibly akathisia) with two more patients requiring dose-reduction due to the same anxiety symptoms, representing an EPS rate of 30%. In another NDA-evidence study, Bright-Asare (1980) showed that 6 of the 20 MCP treated GERD patients suffered EPS, with three dropping out of the study due to this toxicity and one developing torticollis that required treatment. These 30% EPS incidence rates far exceeds the 1/500 rate that made it into the product label.

Additionally, two unpublished NDA studies by Cohen and Sturdevant both provided MCP-EPS rate data of 53% (9/17 reported restlessness or anxiety) including one dropout. Therefore, in the NDA studies I reviewed 5/64 (8%) MCP patients dropped out due to EPS toxicity that they considered intolerable (compared to none in the 52 placebo controlled patients). Eight percent 42.2% or 53% are all far greater than the labeled 0.2%!

A summary of the safety data for the GERD indication are on the table below:
TABLE

As a result, the oral tablets were offered for wide marketing and were available to be prescribed in the U.S. with the misinformation that EPS (all forms) was rare, and mainly self-limiting acute dystonia. Pall and Williams (1987) warned that inappropriate prescribing of metoclopramide for chronic nausea was resulting in types of EPS, including TD, that were consequential.

Newton-John (1988) reported a higher incidence of EPS to high dose MCP in patients aged 13 -- 29 years (27,3% vs. 1.8%) in persons 30 -- 72 years old. Bateman et.al., (1989) prospectively examined the incidence of EPS amongst patients receiving their first MCP prescription (May to November 1986) in a region of the United Kingdom. Most, 1330/2170 or 61% (where indication was known) were for nausea and vomiting, with an additional 624/2170 or 29% for dyspepsia). Females comprised 65% of the patients. The median dose was 30mg daily. They received 12 reports of acute dystonic-dyskinetic events from the physicians of the 2557 exposed for an overall rate of 1/213 (95%CI: 1/136 -- 1/489). The rate was greater (1/81) for this short-term use in those under 30 vs. older than 30 years (1/572), which confirmed the signal that arose in the spontaneous reports within the UK yellow card scheme. There were also 5 reports of Parkinsonian reactions with an overall incidence of 1/512 (age specific rate of 1/275), all in persons over 40 years. In addition, there were 8 occurrences of akathisia with no clear age-related pattern. Therefore the overall detected and reported EPS rate was 25/2557 or 1/102. The specific duration of use was not provided. Note that this study required that the ADRs be reported to the physicians by the patients (who were not aware of the surveillance), therefore they were likely to underestimate the true rates. Even so, these EPS rates were far greater than the labeled, 1/500. The authors were critical of the use of MCP for N&V in the young given the 1/81 chance of an acute dystonic reaction. These studies provided prime evidence for differing age and gender specific rates for the sub-types of EPS induced with non-long-term MCP exposure.

Also the same year, Miller and Jankovic (1989) published a comprehensive review of the subject of MCP-induced movement disorders. They reviewed the 16 patients (10/16 had TD; 15/16 were continued on MCP after the onset of EPS, likely due to non-attribution) with serious MCP-EPS (Note: acute and self-limiting EPS would not present to this clinical site) from their movement disorder referral center. They also reviewed the literature and noted 1031 cases of MCP-TD. The authors question the 1/500 (0.2%) rate and suspect it to be orders of magnitude more common. This one report, if alone, should have been enough to support a change in the MCP label to remove the misleading 1-in-500 rate and to focus on the problem of MCP-TD due to unapproved long-term therapy!

Robinson et.al., (1994) reported a variety of acute EPS symptoms in a 76-year-old diabetic male on MCP 10mg twice daily that abated 36 hours after discontinuation. He warned that diabetic persons being treated for GERD might be at higher risk for EPS. van Harten (1997) found strong positive correlations between the hyperkinetic subtypes of EPS (i.e., tardive dyskinesia, tardive dystonia and akathisia) when they examined a population of psychiatric inpatients. They found that 30% of patients suffered from 2 or more forms of EPS. For example, having drug-induced TD increases the likelihood of developing akathisia six-fold. Parkinsonism was inversely related to TD. These results show that the hyperkinetic subtypes have similar etiologies.
Acute dystonia

The EPS sub-type of acute dystonia appears to have an age relationship that contrasts markedly with both Parkinsonian and TD forms of EPS. Acute dystonic reactions are typified by rapid onset (24-72 hours after initiating therapy) and dissipates within a few hours following exposure.

Two early publications by White and Kennedy (1981) and Gupte (1981) document acute dystonia following normal IV dosages of MCP for nausea and vomiting in children. White and Kennedy found EPS reactions in 33% of treated children. Gupta reported on 18 cases. Kris et.al., (1983) documented an age differentiation in the risk for acute dystonia following high dose IV MCP, with the highest rate (27.3%) in the under 30 group.

Bateman (1989) observed and overall incidence rate (1/213) for acute dyskinesia given short-term use using the UK's general practitioner research system to follow-up on the spontaneous reports received by the UK's Committee on the Safety of Medicines. They examined 2557 patients who received their first prescription most for nausea and vomiting or dyspepsia) for oral MCP and found 12 cases (8 were in patients under 30 years) of dystonic-dyskinetic events. Those under 30 had a far higher (1/81) than the rate in older persons with the same use pattern. These EPS rates in the young were far greater than the labeled, 1/500. Twenty-four children (10 under 6 months of age) presented to the ER with CNS toxicity induced by MCP during one year (1991) of observation, 19 had typical oculogyric crisises and 13 had dystonia (Al-Zaben, 1995). Most were prescribed MCP for vomiting. The authors considered MCP use as unnecessary. Cezard et.al., (2003) report on 81 cases of acute dystonia in children. Reports of 184 cases of MCP poisoning to poison control centers (Note: In France, these are also used as pharmacovigilance collection centers) from 1995-2000 were analyzed retrospectively. Consistent with case reports, they could find no dose-effect correlation for acute dystonia.

Kerr (1996) reported two cases of acute dystonia in adolescents. Batts and Munter (1998) report on a six-month old accidently overdosed (6 fold) with 24mg over nine hours who developed dystonia. A ten year-old boy experienced supraglottic dystonia induced by MCP in a report by Talt (2001). Gokhale (2004) shows that through the placental transfer of MCP from a single maternal dose, a newborn developed dystonia. Yis et.al., (2005) report two dystonia cases in children acutely treated with usual doses of MCP for chemotherapy-induced emesis. Both abated within an hour after diphenhydramine injections. They note the difficulty in diagnosing acute dystonia in a child who is distressed and vomiting.
Akathisia

Akathisia (from the Greek for “not to sit”) is an EPS sub-type consisting of a subjective feeling of needing to move, which is often manifested in an inability to sit still. This is usually in addition to or followed by the objective motor signs of restlessness. It is frequently mistaken for the onset or an exacerbation of psychotic symptoms, anxiety and/or depression.

Neuroleptic-induced akathisia (NIA) is a common side effect with CAPs. Its prevalence is estimated to be 20% (Adler 1989). NIA is classically thought to occur acutely and result in leg movements that are semi-voluntary.

Jungman and Schoffling (1982) found that 25% of persons receiving an MCP IV-bolus in test conditions complained of akathisia that lasted for 3-4 hours. A middle-aged male on normal doses of MCP for 6 weeks for GERD was reported by Hamilton (1987) to have developed Akathisia. Miller and Jankovic (1989) found 10% of the 1,031 reported MCP-induced movement disorders had akathisia.

Tsai (1996) saw akathisia in a 72 year-old male after long-term exposure that was successfully dechallenged. A case of severe MCP-induced akathisia led to a suicide attempt in a 19 year-old was reported by Chow (1997). LaGorio et.al., (1998) illustrated two cases that developed akathisia soon after receiving pre-op doses of 10mg MCP IV, resulting in a need to cancel surgery. Patterson et.al.,(1999) investigated the efficacy of MCP in the treatment of diabetic patients with symptoms of diabetic gastroparesis (DG). They found modest efficacy with 40mg orally up to 4 weeks but an extremely high rate of CNS toxicity, mostly somnolence. Akathisia was considered an adverse event in about 35% of the MCP treated patients, a rate that was greater than the labeled, 1/500 or 0.2%.

Poortinga and colleagues (2001) diagnosed akathisia in a pregnant woman taking oral MCP for nausea, which was first thought to be anxiety or panic disorder. Anfinson (2002) found that 2 days of dopaminergic blockage due to MCP produced akathisia, panic disorder, agoraphobia, and major depressive disorder which all developed sequentially leading to months of disability. The RATE of MCP infusion was shown to affect the severity and incidence of akathisia in a controlled randomized clinical trial by Parlak et.al., (2007). They compared slow infusion to a bolus dose and found 5.8% and 28.8% rates respectively. These EPS were far (up to 144 [28.8% / 0.2%] times!) greater than the labeled, 1/500 or 0.2% rate.
Parkinsonism

Drug-induced Parkinsonism is identical to idiopathic Parkinsonism in its clinical presentation and is characterized by the classic triad of tremor, rigidity and bradykinesia (i.e., slow movements).

Indo and Ando (1982) identified MCP as a major cause of drug-induced Parkinsonism. They reported on the clinical characteristics of 10 cases from Japan. All received normal 30mg/day doses but duration was extended, often beyond the onset of symptoms. There were seven females with most over 60 years. After diagnosis and discontinuation of MCP, it took and average of 4.2 months to reverse the Parkinsonism.

Buchholz and Kariya (1983) describe a case of a 33 year-old diabetic treated with MCP 10-40mg QID for gastroparesis and diminished kidney function who developed Parkinsonism that was positively dechallenged. Sirota et.al., (1986) reported two cases of MCP-induced Parkinsonism in persons with renal disease whose symptoms were positively dechallenged. Maricle and Leung (1989) described 66 year-old female who was prescribed MCP five years earlier for post-operative nausea that was continued in error. She developed Parkinsonism that improved within seven days of discontinuation. Sethi (1989) reported six cases of MCP-PD due to chronic exposure seen that their clinic in a two year period. The symptoms improved upon discontinuation and the authors considered this ADR to be not rare. Bateman (1989) reported 5 reports of Parkinsonian reactions with an overall incidence of 1/512 (age specific rate of 1/275), all in persons over 40 years. This rate, however, used the denominator of all users, not correctly restricted to proper chronic exposed user. This Parkinsonism EPS rate, even when calculated conservatively, is greater than the labeled overall rate of 1/500.
Other EPS

In addition to the fours main MCP-induced EPS, many other subtypes have also been observed and diagnosed. Tardive tremor due to MCP was reported by Tarsy & Indorf (2002). MCP-induced supersensitivity psychosis was reported by Lu et.al. (2002).

The elimination of MCP is slowed in poor metabolizers of CYP2D6 or in patients taking inhibitors of this isoform (Desta 2001). This may be a reason for increased CNS-ADR susceptibility seen in various persons.

To conclude, a full array of EPS types were detected (but not clearly recognized) early in the development of MCP. Their manifestation is highly contingent upon route of administration, rate of administration, age, gender, mental acuity, and duration of exposure. The safety databases for oral use in DG and GERD were small. Given the preponderance of oral therapy, akathisia appears to be the most common EPS, occurring in 10-30% of the treated population. Acute dystonia occurs frequently in the young, especially if receiving a rapid infusion of MCP.
XII. Tardive Dyskinesia Incidence and Prevalence

Schizophrenia (i.e., a mental disorder characterized by abnormalities in the perception or expression of reality. It most commonly manifests as auditory hallucinations, paranoid or bizarre delusions or disorganized speech and thinking in the context of significant social or occupational dysfunction) is the most prevalent psychotic disorder and it is amenable to non-curative treatment with antipsychotic drugs (CAPs and ATAPs). It has an onset in the late-teens through the 20s with a 1:1 gender ratio and a lifetime prevalence of 4/1000 (after age 18) with a 7.2/1000 lifetime morbid risk.

The most critical side effect to conventional antipsychotic pharmacotherapy is Tardive Dyskinesia (TD). TD has an incidence (for mid-aged adults) of 5% per year with sufficient dose and the long-term exposure to CAPs such as those required in schizophrenia (Kane, 1982). However, the point prevalence of TD in psychotic patients treated with LT CAPs is around 15 - 24%, reaching 42 - 49% in the elderly. (Bourgois M et al., 1980; Saltz BL et al., 1989). van Harten (1996) found a prevalence of 40% for TD in chronically institutionalized psychiatric population. Cit is important to recognize that cumulative incidence is not equal to prevalence since much of the early drug-induced TD is reversible. Approximately 60 to 70 percent of the TD cases are mild, many reversible, with about 3 percent being extremely severe. Severe cases may involve problems such as difficulty swallowing, speech interference, cosmetic disfiguration, and respiratory trouble. As evidenced below, he elderly are more susceptible to persistent and irreversible Tardive Dyskinesia than younger people. It is the severe cases of TD that are referred to movement disorder clinics. I will testify that obviously, since length of therapy is a major risk factor, elderly patients have CAP-TD rates at the top of the ranges given above.

The onset and severity of CAP associated TD can be altered and minimized by careful monitoring and dose/duration modification. The American Psychiatric Association (hereafter, APA) (Baldessarini,1980; Kane, 1992) has developed numerous guidance documents, which were widely disseminated to their membership, on the best practice required to minimize the risk of TD and especially to prevent or delay the tragic irreversible onset. Even with guidance and training from the APA it took many years to become the standard of care. Based on the material that I have sought and reviewed, I can find no similar documents or training efforts by the gastroenterologist professional societies.

Table 16 from the 1980 Report, entitled “Suggested Guidelines for Avoidance and Management of Tardive Dyskinesia”, is reproduced here:

1. Consider indications for prolonged neuroleptic therapy carefully; indications chronic psychosis) should be serious, with objective evidence of benefit.

2. Seek alternative therapies in neuroses and mood and character disorders.

3. Use lower doses in elderly patients and children, strive for minimum effective doses, avoid multiple drugs, and remove antiparkinsonism agents as soon as possible.

4. Advise patients and families of risks and benefits; arrive at a mutual decision when use of neuroleptic exceeds one year. Note discussion and agreement in the clinical record.

5. Examine patient regularly for early signs of choreoathetosis and oral-lingual dyskinesia. Consider alternative neurologic diagnoses.

6. Reevaluate and document indications and response at least every three to six months and attempt to reduce dose.

7. At earliest sign of dyskinesia, lower dose, change to a less potent agent, or ideally stop treatment; await remission as long as psychiatric status permits.

8. Treat dyskinesia with benign agents first (diazepam, deanol, choline, or lecithin in high doses, possibly lithium); stay alert to new experimental therapies, if only to bide time and offer hope. Reinstitute neuroleptics only as an extreme measure for disabling dyskinesia, using lowest doses feasible.

Atypical antipsychotic agents (hereafter, ATAPs) were developed because they have a lower risk for TD and have largely replaced CAPs as first-line therapy for schizophrenia. ATAPs are, however, associated with TD although they have been shown to have a lower (maybe 25% that of CAPs) risk for TD than CAPs. CAPS (especially Haloperidol) are still used to treat behavioral problems in the mentally retarded and in the aged demented populations.
Age, gender and sub-population specific rates

There are many risk factors associated with the occurrence of dyskinetic movements including age, gender and length of neuroleptic (including MCP) exposure. Bourgois et.al., (1980) found that the incidence of dyskinesia in elderly females (27%) was more than twice that of elderly males (12%). Kenney (2008) found that 77% of confirmed severe TD patients were female vs. only 50% for other movement disorders referred to their clinic.
Spontaneous Dyskinesia

Spontaneous Dyskinesia is reported in adults, but most often in elderly persons. Bourgeois et.al., (1980) conducted a comparative study of 270 elderly French subjects in a retirement facility, where about 3/4 of residents had never been treated with antipsychotics. There were 191 (71%) females and 79 (29%) males in the study of which 1/4 (N=63) had movement disorders. The average age of those with movement disorders was 81.8 vs. 76.1 for those without. The females experienced twice the risk for dyskinesia, 27% vs. 12%, than the males. Those persons who had been exposed to neuroleptics (likely CAPs) had a risk of 2.25 times (42 vs. 18%) those who were not exposed. Thus they found an 18% rate of spontaneous dyskinesia in a very elderly population. The authors acknowledged that the non-exposed status was based on the medical records in that facility and noted history, so it may be possible that earlier exposure was not reported. The gender specific rates for spontaneous dyskinesia were not provided.

Yassa and Jeste (1992) assessed five published studies and found a mean prevalence of spontaneous dyskinesia to be 13.9% (7.4% in males; 18% in females). These data covered a wide age range. The point prevalence of spontaneous varies and increases with age with best adult estimate being 5% (range 3 -77%) (Tarsy & Baldessarini 2006).

These studies demonstrate that the spontaneous dyskinesia rates found in the elderly VA populations studies by Ganzini (1993) and Sewell (1994) were not contrary to what could be expected.
XIII. Off-Label Long-Term Use of Metoclopramide

There was an understanding by the FDA in the early 1980s that MCP, a dopaminergic agent, had the potential for TD given LT exposure. Initially it's use was limited to inpatients and severe DG. However, it was eventually approved for up to 12 weeks treatment of GERD.

The reports of MCP-TD presented in Section X of this report all report TD as a consequences of off-label long-term use. Grimes (1982b) and Wiholm (1984) were the first to analyze the signal and present the problem.

Miller and Jankovic (1989) found the TD emerges, on average, 12 months of exposure and that treatment was continued for an average of 6 months AFTER involuntary movements had started. Sewell (1992a) reviewed the reports in the literature and found 67 cases of MCP-TD with 20 months duration of exposure.

Stewart (1992b) analyzed “inappropriate” long-term use of MCP in the elderly. This descriptive study identified 34 exposed to MCP from 4515 (2898 females; 1617 males) patients who participated in a health screening research program in Florida, 11 (32.4%) were long-term users. The study was initiated because two patients presented to the University hospital with MCP-induced TD within the prior year. Both had taken MCP for longer than 8 years for GERD. Only three of the 34 participants had DM, so GERD or heart-burn-like ailments was their likely treatment diagnosis. This small study signaled that LT-MCP was a substantial component of MCP use in the elderly.

These published cases and the studies of Stewart, Ganzini, and Sewell made the problem of long-term off-label use visible and highlighted its consequences.

To assess risk factors for TD in MCP exposed persons, Schaffer et.al. (2004) reviewed the reports in the FDA Adverse Reaction Reporting System (AERS) through June 2003. They also used the IMS Health data to evaluate MCP prescribing patterns before and after the withdrawal of cisapride from the US marketplace. They extracted a case-series of 87 reports from AERS with TD as the adverse outcome and MCP as an exposure. Each case was re-evaluation for validity of both exposure and outcome. Data regarding prescription utilization (via National Prescription Audit, NPA) and demographics of patients (National Disease & Therapeutic Index, NDTI) treated with MCP, between 1/1992 - 6/2003, were obtained from IMS Health. The mean age for the case-series was 60 years and 2/3 were female. The mean dose was 33 mg/day with mean duration of 753 days (~108 weeks). Most of the known use was for GERD (gastroesophageal reflux disease), or acid reflux, whose main symptom is heartburn. Only 14% of use was for gastroparesis.

The marketing of cisapride in 1993 had an inverse impact on the sales of MCP, which reversed when cisapride was withdrawn from the market due to QT wave prolongation. Nearly 1/4 of MCP use was in persons, mostly females, over 70 years of age. This descriptive analysis reinforced the fact that a meaningful percentage (15-20%) of MCP is used long-term (off-label) in persons (elderly females) who are at the highest risk of TD. The authors reflect on the impact of these prescribing practices given the ‘paucity of evidence that MCP improves the quality of life” and the risk of a life- changing, potentially irreversible neurological disorder.

Kaplan (2007) used medical claims data to access the proportion of MCP use that occurred beyond the 12 week maximum that was evaluated in the pre-marketing clinical trials and appears in the approved product labeling. They abstracted prescription data (from 2002 -- 2004) from the Caremark longitudinal patient-level database and calculated the duration of MCP exposure from the 200,907 participants (most all were adults with mean age of 58; 2/3 female) who had prescriptions of MCP. They found that 15% of users had exposure that exceeded 90 days duration with cumulative therapy of 90 days or more in almost 20%. The extended users were also 2/3 female but with an average age of 65 years, a demographic cohort that matches those at highest risk for TD.

These data illustrate that despite a current label with TD included, prescribers continue to place patients at risk for potentially persistent and disabling TD. The authors did not have access to indication data so the underlying diseases being treated with MCP could not be analyzed. The number of persons at risk has increased after the market withdrawal of cisapride in 2000, and the subsequent rise in MCP use (Schaffer, 2004).
XIV. Incidence & Prevalence of MCP-Induced Tardive Dyskinesia

No incidence studies of MCP-induced TD have ever been conducted by any NDA or ANDA holder. The incidence of MCP-induced Tardive Dyskinesia in any sub-population is unknown and cannot be directly calculated from existing data. The estimates in the label remain unclear (although they are likely from the x paper and pertain to single dose inpatient risk) in origin and do not reflect those that occur in current clinical practice, for MCP-TD or any resultant EPS subtype.

When MCP was first marketed TD was presumed rare given MCP's pharmacology and short-term use indications. However, isolated case reports were published in the literature shortly after its marketing. As discussed above, Wiholm (1984) presented 11 MCP-induced TD cases from Sweden, most in elderly females upon long-term exposure. He calculated reporting rates and estimated the risk (given long term exposure) to be no less than 1:1000. This was an important estimation given the initial presumption of extremely low or negligible risk. However, Wiholm's data, which was from the Swedish spontaneous reporting system, should never be used to calculate true incidence or prevalence estimates. They can, however, signal that something is amiss. That is, an outcome that was previously only associated with long-term use of neuroleptics was occurring with a drug only approved for short-term use and was not used at a psychotropic. This signal should have been pursued and verified by the Applicant.

When both the off-label long-term use of MCP and it's tragic neurological consequences became generally known it was independent researchers and not the Applicants who attempted to assess the magnitude of the problem. Below is information on three of those independent efforts.

Ganzini et.al. (1993) conducted a cross-sectional study with the stated goals of 1] to determine the prevalence and severity of TD and acute EPS in MCP-treated patients and 2] to compare the prevalence and severity of TD in MCP-treated diabetics and non-diabetics. Fifty-one persons from a group of medically ill Veterans who were prescribed MCP between May and August 1991 met the inclusion criteria and agreed to participate.

Controls were selected from the appointment lists of the same VA Medical Center and matched on gender, age and presence of diabetes and absence of dopamine agonist exposure and diagnosed movement disorders. For this study persons treated with insulin or oral hypoglycemics were assumed to have DM. Persons not on those agents but having fasting plasma glucose above set levels were also considered diabetic. The Abnormal Involuntary Movement Scale (AIMS) was used to measure TD. Control subjects who met the set definition of TD were considered to have spontaneous dyskinesia. The authors acknowledged the following factors as risk factors (from cited literature) for TD: increasing age, female gender, an affective disorder, anti-cholinergic medication use, and the increasing dose and duration of neuroleptic treatment. They cited Ganzini (1991) as having raised suspicion about diabetes as a TD risk factor. They were aware that only severe cases were reported in the literature or are referred to movement disorder clinics, so the true prevalence could only be derived from a cross-sectional cohort or a prospective cohort study. They quote the MCP product label's EPS incidence of 0.2% (1/500) but wondered upon what evidence base those estimates were based and if they had any relevance to (then) current usage patterns of MCP. That because their clinical experience suggested that the rates are “substantially higher”.

The results after a blinded evaluation of the male (50/51) Veterans gave an average age of 63.6, dose of 31mg/day and a mean duration of 2.6 years. Fifteen (29%) of the 51 MCP users met the case definition for tardive dyskinesia. In the Veteran controls of similar age and gender, the prevalence of spontaneous dyskinesia was 17.6%, providing 12.6% excess in the treated group. The MCP-TD was more severe than the spontaneous dyskinesias at a statistically significant level. A relative risk point estimate of 1.67 (95% CI: 0.93 -- 2.97), indicated a 67% increased risk but that finding was not statistically significant at conventional levels. Of interest, is the finding regarding Parkinsons. Thirty-one percent (n=16) of MCP users met the case definition of drug-induced Parkinsonism versus 7.8% (n=4) in the control cohort for a RR=4 (95% CI: 1.5 -- 10.5). Akathisia was more prevalent in the treated group (N=6; 13.7%) versus the controls (N=3; 5.8%) but due to small numbers there was no statistical significance. Consistent with the long-term exposure there were no cases of dystonia in the MCP exposed. These data regarding these two EPS are supportive of and internally consistent with the TD finding and EPS is commonly prevalent in this treated population. Ganzini states that MCP-EPS in this treated population may be 100 times greater than reported in the product label. This is an extremely strong statement, true or not, that should have alerted any relevant Applicant.

The author further relate their results to those seen with neuroleptics, in the aged medically ill and as it pertains to spontaneous dyskinesias. They find the consistency that implies validity. Lastly they discuss the lack of clinical recognition of these outcomes in non-psychiatric cohorts and support their finding with literature citations.

Sewell et.al. (1994) studied the frequency of dyskinesia in patients with a history of exposure to MCP compared to a non MCP-treated comparison group. They also studied Veterans (mean age 60; 80-90% male; 44% with DM) from another VA Medical Center. Identified patients had a history of MCP greater than 30 days, no exposure to another neuroleptic, absence of current DSM-II substance abuse disorders and the absence of a pre-existing movement disorder. A psychiatrist using a structured clinical instrument interviewed patients and their medical and pharmacy records were abstracted. Blinded' raters assessed EPS using validated instruments (modified AIMS). The presence of TD was clinically defined using published criteria except that the required length of exposure was one month (instead of 3 months) because of the increased susceptibility of older patients. Dyskinesia was confirmed in 14/51 (27.4%) MCP-exposed and 4/34 (11.8%) comparison subjects. Most were mild, 3 were of moderate severity and no were severe. Diabetes existed in 9/14 (64.3%) MCP-exposed TD subjects and 14/37 (37.8%) non-TD subjects, a statistically significant difference. Five of the 14 patients were followed for an additional 10 months. In the four where the MCP was discontinued, the TD persisted at 11 months. The authors conclude that “MCP-associated TD may be frequent, mild to moderate in severity, and persistent”. There was a 27.4% period prevalence in the MCP-exposed group, an 11.8% prevalence of spontaneous dyskinesia in the non-exposed group, and a risk difference of 15.6%. Diabetes was confirmed as a risk factor of TD. They note that unlike the major psychiatric illnesses for which CAP or ATAPS are used, MCP can be more readily discontinued because alternate treatments exist and the clinical consequences of discontinuation are less severe.

Matson et.al., (2002) point out the real risk of MCP-TD afforded specific sub-populations. One of the reasons for the study was to determine if there was an elevated risk of TD in persons with mental retardation, given that they are often treated long-term of gastrointestinal disturbances. They examined information on 75 adult (average age ~ 40 years) persons with mental retardation that reside in a facility in Louisiana. Subjects were equally classified into three groups based on type of medication administered: metoclopramide (average 33mg/day), conventional antipsychotics (CAPS), and neither. They were matched on an individual basis on age, gender and level of mental retardation. Medical records were abstracted for information regarding DSM Axis-I diagnoses and other psychotropic medications. Available records confirmed that none of the 25 persons in the MCP group had a history of exposure to any other dopamine-blocking agent. All MCP exposure was at least 4 years of duration. Participants were assessed using a validated (including specific validation of the mentally retarded population) 15-item TD rating scale called DISCUS as part of a routine quarterly examination of medication side effects. Matson retrospectively reviewed these data. The analyses of trends or differences amongst groups in this study were hampered by small sample size. Ten of the 25 patients or 40% in both the MCP and CAP groups met the definition for probable TD during the year examined. Two (8%) persons in the MCP group and five (20%) of the CAP group were rated as having persistent TD, given long-term exposure. As a probable result of close monitoring and dose-modification, the average level of TD severity was low and trended down during the monitored year. However, it showed that 40% of long-term users in this population had symptoms of TD and 8% had persistent disease. These are annual prevalence figures.

The point prevalence for of LT-MCP induced TD is estimated at 13-40% with MCP-LTT is taken from these three published studies (Ganzini (1993), Sewell (1994) and Matson (2004). In another MCP case, Wyeth without notice at my deposition, asked me to estimate the incidence rate of TD. The subsequently court found my logic acceptable, and I estimated the rate as follows: “In the context of the published literature on neuroleptic (CAP) induced TD and the findings of Wiholm and others and in the absence of any Applicant supported and properly conducted research, I used the two published studies in elderly males to address whether LT-MCP-TD in the elderly is rare or not. Using the difference in the point prevalence estimates of 11% from Ganzini (1993) and 15% from Sewell (1994) -- and with Matson (2002) in mind -- I used an unweighted average of 13% prevalence figure to represent those cases that are most likely induced by MCP. Then I used three-year duration of exposure to estimate that it would take about a 5% incidence in each of those exposure years to reach a true prevalence of 13%”. A 5% incidence would be a 1/20 estimate and that would indicate that TD is common in LT-MCP exposed elderly sub-population. As noted, these data are from male patients and the rates in females may be substantially higher. Yet, the product label, had read and still reads that the risk, for all EPS symptoms combined is 1 in 500.

Kenney et.al., (2008) examined the main causes of TD in a university-based movement disorders clinic. They reviewed and validated 434 medical records of patients referred to the Center between 1981-2006, who met the accepted clinical definition of TD. Females numbered 334/434 or 77% of cases. This is in contrast to 50-50 gender balances for all other movement disorders, including Parkinsonism, seen in their clinic. The average age was 64 years. Treatment indications of all 434 patients were psychiatric (68.2%), gastrointestinal (30%) and other (1.8%). The most common medications associated with the onset of TD were haloperidol (N=191; 44%), MCP (N=171; 39.4%), amitriptyline - perphenazine (N=85; 19.6%), and thioridizine (N=72; 16.6%). The most common cause from 1981 -- 1999 was Halperidol. Importantly, from 2000 -- 2006, MCP-induced TD was more common than any other cause, accounting for 34.5% (87/250) of all cases! The authors suggest that, if clinical alternatives exist, avoiding DRBDs is the best approach to minimizing the development of TD, especially in the elderly female. The next best is to be able to recognize early sign and symptoms and initiate appropriate treatment, including discontinuation of the offending agent. They state that “patients treated with (long-term) MCP require close follow-up and careful examination for stereotypy, dystonia, and chorea as recommended by the American Psychiatric Association for patients treated with either first- (i.e., CAPs) or second-generation (i.e., ATAPs) neuroleptics”.

As discussed above and in my prior deposition (Kettering) testimony and as reported in the literature, the 1/500 occurrence in the product label is NOT accurate for TD in any (short or long term, young or old) treated population. The 1/500 figure was an estimate developed for single-dose inpatient IV use (and the Bodi (1966) study supports a 1 in 8 EPS rate for oral dosing). It is fundamentally unsupportable and to the contrary of all empirical evidence developed and available (though ignored by the Applicants). It is misleading and inaccurate to apply this 1/500 rate to any EPS events resulting from specific uses &/or exposed populations, for example:

• The risk for acute dystonias in patients (most notably in young adults <30 years and children) treated with high doses for prevention of N&V secondary to CA chemotherapy is orders of magnitude higher.

• The risk for TD given short exposure to MCP is rare in adults and approaches zero in the non-elderly adult. Therefore, any TD risk seen with long-term therapy is an increased risk.

• The TD risk in the mentally challenged, diabetics or the elderly may be increased over other adult cohorts, given LT exposure.

Therefore, no single risk or rate estimate is sufficient to describe MCP-induced TD in clinical practice sub-populations.

Though approved by FDA only for short-term use, MCP is used for the chronic treatment of diabetic gastroparesis and GERD. That long-term exposure places patients at risk for TD. At present, MCP-induced TD was more common than any other cause, accounting for 34.5% (87/250) of all severe cases referred to movement disorder centers! This is a surprising and alarming public health finding! Kenney (2008) indicates that the MCP-TD problem remains and severe TD still occurs, now at alarming rates, because gastroenterologists are still not diagnosing and mitigating. This problem demands action!

The incidence of MCP-TD can only be calculated by determining the number of new cases in a cohort of persons exposed to MCP for greater than 3-6 months. To be of most communication or information value, incidence rates should be calculated for a number of discrete and specific cohorts (i.e., mid-aged adult; elderly; retarded; diabetic) since evidence suggested that they are likely to differ. Two studies were conducted in focused population (here, Veterans Administration patients) in hopes of assessing the commonality of TD associated with long-term MCP exposure. The first was a cross-sectional study by Ganzini (1993), which both documented long-term exposure (M = 2.6 years) and confirmed the increased presence of TD in those persons exposed to MCP (Point prevalence: 29% in the exposed vs. 18% in the non-exposed with a difference of 11%). They also found a four- fold difference (31% vs. 8%) in MCP-induced Parkinson's symptoms. A vast under-recognition of MCP-TD was documented and whether true or not, the 100X excess vs. current label statement they made should have, at least, catalyzed a comprehensive risk assessment by the Applicant. The Applicant apparently paid this study little regard.

The second study was also a cross-sectional design conducted by Sewell (1994). They further documented long-term exposure and further confirmed the increased presence of TD in those persons exposed to MCP (Point prevalence: 27% vs. 12%, with a difference of 15%). Although conducted in older males, I am of the opinion that both the Ganzini and the Sewell study results are generalizable to females of similar age. They are, of course, NOT generalizable to all MCP users, since most are NOT long-term users.

Both these studies identified mild-moderate cases, which are likely to be more prevalent than severe cases depending on presentation to healthcare, diagnosis and proper medical actions beginning with removal of the offending agent. Avorn (1995) study on Parkinsonian symptoms illustrates gastroenterologists' general ignorance of MCP CNS ADRs, which aids the explanation of why most of the Ganzini and Sewell TD patients were not previously diagnosed. These studies evidence that that the likelihood of an elderly person having developed TD with LT-MCP may be common, and NOT rare, as implied by PLIVA's label for metoclopramide. Yet, PLIVA has not investigated this literature to ensure MCP's safe use.

a- High Risk Sub-groups for TD

There are many risk factors associated with the occurrence of dyskinetic movements including DM, age, gender and length of neuroleptic (including MCP) exposure.
Diabetes Mellitus

The relationship between DM & TD has long been suspected. Two studies were performed by Ganzini (1992) to assess the relationship between TD and DM. In the first, they compared two groups of 38 persons each, one with DM, the other not, who were chronically treated with a steady dose of a CAP. The elderly men with DM had a 79% prevalence (30/38) versus 53% (20/38), a statistically significant difference. The TD was also more severe. The second study looked specifically at MCP (also a neuroleptic by pharmacology) treated patients. They selected 51 subjects from the Veterans Administration population who took at least 20mg/day (average: 31mg) for 3 months (average: 2.6 years) duration. They compared the prevalence of TD in the 24 diabetics and the 27 non-diabetics and found that TD was found in 42% (10/24) of the diabetics and 19% (5/27) of the non-diabetics (p = 0.07).

Sewell et.al., (1992) present two case reports in hemodialysis patients with DM who developed MCP-induced TD within 12 months of beginning treatment for DG. They discuss that non-psychiatrists ware less likely to detect and diagnose early TD.

Woerner (1993) examined 160 elderly patients at the initiation of CAP treatment. After 43 weeks of exposure to neuroleptics, the cumulative incidence rates from TD were 54.1% for the patients with diabetes versus 25.6% for those who were not. This provided a risk ratio of 2.38 (95% CI: 1.04 -- 5.46). When the analysis was limited to only those on drug at the time of the TD diagnoses, the rate ratio increases to 4.14 (95% Cl:1.04-16.50).

Acute diabetic gastroparesis (DG) is an approved indication for MCP. DG can be a chronic condition. Chronic treatment of DG is an off-label use of MCP. It is the long-term treatment of DG (and GERD) with MCP that places patients at risk for TD.
Gender

Bourgois et.al., (1980) found that the incidence of dyskinesia in elderly females (27%) was more than twice that of elderly males (12%). Yassa and Jeste (1992) conducted a critical review of the literature on gender and TD. They assessed 76 selected published studies through 1989 and found an overall prevalence of 24.2% in CAP treated patients. The prevalence was significantly higher (26.6%) in females when compared to males (21.6%). Peak overall prevalence was in the 50-70 year-old age group, but continued to increase in females to peak at 49.1 %. Females also tended to have more severe TD. Kenney (2008) found that 77% of confirmed severe TD patients were female vs. only 50% for other movement disorders referred to their clinic.
Mentally Challenged

Matson et.al., (2002) point out the real risk of MCP-TD afforded specific sub-populations. One of the reasons for the study was to determine if there was an elevated risk of TD in persons with mental retardation, given that they are often treated long-term of gastrointestinal disturbances. They examined information on 75 adult (average age ~ 40 years) persons with mental retardation that reside in a facility in Louisiana. Subjects were equally classified into three groups based on type of medication administered: metoclopramide (average 33mg/day), conventional antipsychotics (CAPS), and neither. They were matched on an individual basis on age, gender and level of mental retardation. Medical records were abstracted for information regarding DSM Axis-I diagnoses and other psychotropic medications. Available records confirmed that none of the 25 persons in the MCP group had a history of exposure to any other dopamine-blocking agent. All MCP exposure was at least 4 years of duration. Participants were assessed using a validated (including specific validation of the mentally retarded population) 15-item TD rating scale called DISCUS as part of a routine quarterly examination of medication side effects. Matson retrospectively reviewed these data. The analyses of trends or differences amongst groups in this study were hampered by small sample size. Ten of the 25 patients or 40% in both the MCP and CAP groups met the definition for probable TD during the year examined. Two (8%) persons in the MCP group and five (20%) of the CAP group were rated as having persistent TD, given long-term exposure. As a probable result of close monitoring and dose-modification, the average level of TD severity was low and trended down during the monitored year. However, it showed that 40% of long-term users in this population had symptoms of TD and 8% had persistent disease. These are annual prevalence figures.
XV. Product Labeling

The 12/14/1976 draft labeling (Bates 1020317) for the oral product stated the following: “Parkinsonism and/or other Extrapyramidal symptoms have been reported in approximately 1 percent of patients. They consist most often as a felling of restlessness, facial grimacing, involuntary movement, oculogyric crisis, rhythmic protrusion of tongue or trismus. Such reactions appear to occur more frequently in children and young adults, and particularly at higher-than-recommended dosage”. The reference for the 1% incidence figure was Robinson (1973). In the original submission the NDA was not approved due to lack of demonstration of efficacy.

Additional efficacy studies for DG and GERD were commenced for the oral dosage forms, but the Applicant continued forward with an NDA for an IV dosage from. Metoclopramide under the trade name of REGLAN was approved in US on February 7, 1979 for:

“Single doses of metoclopramide may be used to facilitate small bowel intubation (e.g., for biopsies) in adults and children in whom the tube does not pass the pylorus with conventional maneuvers.

Single doses of metoclopramide may be used to stimulate gastric emptying and intestinal transit of barium in cases where delayed emptying interferes with radiological examination of the stomach and/or small intestine”

The original product labeling submission for these single dose IV indications included:

WARNINGS: Metoclopramide can cause extrapyramidal reactions, therefore it should not be used in patients with epilepsy or extrapyramidal syndromes unless its expected benefits outweigh the risk of aggravating these symptoms... In view of the higher risk of Extrapyramidal manifestations, metoclopramide should be used in children only if a clear indication is established.

ADVERSE REACTIONS: The most concerning adverse effects are Extrapyramidal dystonic reactions, most often a feeling of unrealness or restlessness but may include restlessness of the limbs, trismus, rhythmic protrusion of the tongue, spasm of extraocular muscles, a bulbar type of speech, muscular twitching particularly of the face and neck, and neck retraction. A report of a dystonic reaction resembling tetanus has appeared. Extrapyramidal reactions appear to occur more frequently in children and young adults. They are associated with higher than recommended dosage, and in hypersensitive individuals may occur after an initial normal dose. The more overt manifestations of Extrapyramidal symptoms occur in approximately 1 in 500 patients...

Symptoms are self-limiting and usually disappear within 24 hours of withdrawal of metoclopramide.

The approved label (December 1978) for the injectable Reglan included:

“CONTRAINDICATIONS: Metoclopramide should not be used in epileptics or patients receiving other drugs which are likely to cause extrapyramidal reactions, since the frequency and severity of seizures or extrapyramidal reactions may be increased.

WARNINGS: Extrapyramidal symptoms occur in approximately 1 in 500 patients treated with metoclopramide. These occur more frequently in children and young adults and may occur after a single dose. These most often consist of feelings of restlessness, occasionally they include involuntary movements of the limbs and facial grimacing, rarely, torticollis, oculogyroic crisis, rhythmic protrusion of tongue, bulbar type of speech or trismus. One dystonic reaction resembling tetanus has been reported”

My comments on this initial label are:

- It is my opinion that they lumped acute dystonic reactions and akathisia under this description. This is likely because the term akathisia was not widely understood or in use (outside of psychiatry) that that point in time. The first literature citation of its association with MCP was Schearer (1984).

- They altered the 1976 label for the oral dosage form that used the 1 % incidence figure from Robinson (1973) for the 1979 approved IV indication

- Since the original NDA only included small numbers of treated patients, it appears that they obtained the “1 in 500” or 0.2% estimate for EPS from one or both of the following sources. 1] French publication by Derbanne (1967) [as reported by J Jones 2007], were 4 cases of EPS were seen amongst 2540 (0.16%) MCP users in a hospital surveillance scheme, 2] early correspondence from the authors Porter and Jick whose 1977 article reported one dystonic reaction in 758 (0.13%) patients during their large-scale case-control hospital surveillance project.

- The 1 in 500 estimate did refer to acute dystonia, which was the likely reaction that would be seen after a single dose of IV product.

Of note, the 1976 Canadian label for the AH. Robins MCP tablets and syrup contained a rate of 1:100 (Bates 1020320). The Canadian label was consistent with that proposed originally in 1976 for the oral dosage from in the U.S.

This makes all the defendants arguments defending the 1/500 estimate very suspect.

The Porter and Jick (1977) data was submitted in the DG NDA application (Bates 1126859) despite having data from the DG RCT indication a likely EPS rate of 11.4%.

On April 13, 1983, Jick did submit follow-up data to Robins and claims only 3 cases from 1552 inpatient exposures (Bates 030740).

Trisdale's (Director, Domestic Clinical Investigations Department, A.H. Robins Company) February 22, 1979 letter to Robins staff and Investigators (citing Lavy, 1978; Kataria, 1978) warns of EPS with MCP and acknowledged that they were seen in Robin's MCP clinical trials. He states, “these cases (i.e., those reported by Lavy and Kataria) point out that a drug with such potential side effects should not be prescribed for relatively trivial indications...”.

The oral tablet was approved for acute diabetic gastroparesis (DG) in hospitalized patients, on October 14, 1980 contingent on the agreement that it be available only to hospital pharmacies and that a post-marketing study be conducted to characterize the use of oral MCP. In other words, because it's approval for this limited indication was based on only 35 patients, the FDA wanted its use to be limited to inpatient DG and wanted continual evidence that it was only being used in that manner. I could find no evidence that this Phase IV study was conducted. The 1:500 EPS rate statement remained on this label, as it appeared in the IV single use version, even though 11.4% of MCP-treated patients (versus zero in the placebo group) in the controlled trials suffered “restlessness” which was likely akathisia.

Despite Bodi, it appears that the Applicant did not elucidate the need to examine EPS during the oral product development for DG and GERD. For example, restlessness and/or anxiety seen in the RCTs were NOT linked to akathisia, even though that would have been prudent given MCP's dopamine-antagonist pharmacology.

MCP was approved for treatment of GERD in 1984 based on a controlled trial safety database of 65 MCP patients with a likely EPS rate of 42.5% with 8% EPS dropout rate. The label retained the 1/5000 rate despite all evidence to the contrary.

As discussed above and in my prior deposition testimony and as reported in the literature, the 1/500 occurrence in the product label is NOT accurate for TD in any (short or long term, young or old) treated population. The 1/500 figure was clearly intended to estimate the EPS risk for the IV single dose product used in hospital (the origin of the figure is unknown but suspected to be Derbanne) and The Bodi (1966) study supports a 1 in 8 EPS rate for the oral administration with short-term use. The 1/500 rate is misleading and inaccurate for many specific uses &/or populations, for example:

- The risk for acute dystonias in patients treated with high doses for prevention of N&V secondary to CA chemotherapy can be much higher.

- The risk of EPS, largely acute dystonias, are increased in young adults (<30 years) and children.

- The risk for TD given short exposure to MCP is rare in adults and approaches zero in the non-elderly adult. Therefore, any TD risk seen with long-term therapy is an increased risk.

- The TD risk in the mentally challenged, diabetics or the elderly may be increased over other adult cohorts.

Therefore, no single risk or rate estimate is sufficient to describe MCP-induced TD in clinical practice sub-populations.

The TD wording was added by R. Lipicky, Director CDER's Division of Cardio-Renal Drug Products (there was no Gastro-lntestinal Reviewing Division at the FDA at that time - 1986) to the MCP label when class labeling was instituted for all neuroleptics (based on data for the CAPs), whether or not they were labeled or used as antipsychotics. As I pointed out previously, at that point in time, there were sufficient published literature reports on MCP-induced TD but no quantitative epidemiological studies. It is my understanding that this label remained the same until 2004.

The current label still does not warn of an increased risk for MCP-TD in persons previously exposed to this or other DRBRs (e.g. CAPs, etc.). The current label still does not warn of an increased risk for LT-MCP-TD in diabetics. The current label still does not warn of an increased risk for LT-MCP-TD in the cognitively impaired. The current label still does not warn that the increased risk for LT-MCP-TD in the elderly is substantial and incidence estimate could be 1/20 or more and common NOT rare. The current label still does not warn that because of an increased risk for MCP-TD with long-term exposure such exposure should be absolutely avoided. The current label still states that all forms of EPS together occur at a rate less than 1/500.
XVI. PLIVA

PLIVA began marketing generic MCP in 1988. I reviewed the depositions of Nicholas Tantillo taken on March 20,2008; deposition of Brian Cooley taken on August 7, 2008 and the deposition of James Morrison in the matter of Linger v. Wyeth. I have also reviewed PLIVA's documents filed in this matter at Pacer documents 27 and 34. I reviewed the documents PLIVA tendered to Plaintiff's counsel in this matter and did not see any comprehensive analyses of this important safety issue. PLIVA is required to submit annual reports to the FDA reporting any safety data and information that might bear upon the safety of metoclopramide. PLIVA is required to review all adverse event reports and conducting postmarketing safety reviews on MCP to determine if there is a safety issue that would warrant a labeling revision. It appears PLIVA did not carry out this mandated task. It appears PLIVA instead relied on Wyeth to review the literature for safety data. All drug manufacturers should be reviewing the literature for safety signals and issues, regardless of whether the manufacturer is a ‘generic’ or innovator. If safety surveillance is not properly carried out under the regulations important safety data could be missed. This is the case with MCP. But PLIVA did not report this literature to the FDA or undertake any assessment of Tardive Dyskinesia, a life altering condition. Instead PLIVA reported that it there was nothing to report. Additionally, confirmatory evidence of causality and for an increased risk for TD with long-term exposure to MCP clearly existed in the literature since 1994 (Ganzini, Sewell) and is consistent in nature with the published evidence (since the mid-1980s) with LT CAP exposure. As discussed above, there were many articles in the literature reporting upon the alarming risk of TD in long-term use. PLIVA did not assess or report any of these findings.
XVII. Conclusions

My opinions on this important public health and patient safety matter are as follows: after reviewing the assembled materials presented in this report, it is clear to me that tens of thousands of individuals may be suffering from a situation that is totally preventable. I am stunned that the applicants, both NDA and ANDA holders, have not rectified the problem over the many years they have been clearly aware of these severe neurologic consequences directly caused by the use, especially the long-term use, of their drug product.

It is my opinion that PLIVA failed to perform a comprehensive risk analysis of EPS including TD.

It is my opinion that metoclopramide was approved for marketing in Europe (1964) before modern scientific standards for efficacy were established and functional.

It is my opinion that metoclopramide (MCP) was approved for marketing in Europe (1964) before even crude pharmacovigilance systems had been established subsequent to the thalidomide disaster in France and Germany in the early 1960s.

It is my opinion that the evidence supports that MCP is a dopamine antagonist and therefore a neuroleptic.

It is my opinion that the Bradford-Hill criteria are an accepted and tested standard for the assessment of a causal relationship between a pharmacotherapeutic agent and an adverse outcome.

It is my opinion that based on the application of the Bradford-Hill criteria short-term MCP exposure is casually associated with the EPS sub-type known as acute dystonia.

It is my opinion that based on the application of the Bradford-Hill criteria MCP exposure is casually associated with the EPS sub-type known as akathisia.

It is my opinion that based on the application of the Bradford-Hill criteria medium- to-long-term MCP exposure is casually associated with the EPS sub-type known as non-idiopathic Parkinsonism.

It is my opinion that based on the application of the Bradford-Hill criteria long-term MCP exposure is casually associated with the EPS sub-type known as tardive dyskinesia.

It is my opinion that the evidence to support the causal relationship between MCP and TD was sufficient as of 1985 and that PLIVA knew about this relationship from the point it's predecessor Sidmak, was sued and notified of metoclopramide related injuries through MCP litigation.

It is my opinion that the evidence supports that the 1/500 EPS rate was a population estimate established for the single-dose inpatient intravenous use of MCP.

It is my opinion that the evidence supports that the Applicant did not revise the EPS section of the product label to reflect the data from the clinical trials performed for the oral dosage form and from the non-U.S. medical literature.

It is my opinion that the 1/500 EPS labeling was inaccurate and misleading from the moment the oral dosage form was approved.

It is my opinion that the evidence supports that acute dystonia is the most frequent (at a rate far higher than 1/500) EPS following acute IV use, especially in infants, children and young adults.

It is my opinion that the evidence supports that Akathisia is the most frequent (at a rate far higher than 1/500) EPS with normal (up to 40mg daily for 8 weeks) dose oral use in adult patients with DG or GERD.

It is my opinion that the evidence supports that older adults exposed to MCP at a normal dose for extended periods are at highest risk for Parkinsonism. Males with decreased kidney function on higher doses are at greatest risk.

It is my opinion that the evidence supports that persons with MCP-induced akathisia are at greater risk for developing TD with continued exposure to MCP.

It is my opinion that the evidence supports that chronic continuous long-term exposure, especially at higher doses, places the patient at risk for TD.

It is my opinion that the evidence supports that there are high-risk groups for MCP-induced TD, including mentally challenged, females, diabetics, medically fragile, and the elderly.

It is my opinion that the evidence supports that TD (any severity) risk rates for elderly females, diabetic or not, with chronic continuous exposure to MCP may be high and relatively common.

It is my opinion that the evidence supports that if recognized early and mitigated (usually just by discontinuation of offending agent), TD is mild and reversible.

It is my opinion that the evidence supports, to the contrary, that if early TD is not detected then mitigated and exposure continues, the risk for severe, debilitating and irreversible TD is increased.

It is my opinion that the evidence supports that the vast majority (approaching 100%) of MCP is prescribed by non-psychiatrists.

It is my opinion that the evidence supports that non-psychiatrists do not have the APA guided training in the detection and mitigation of TD.

It is my opinion that the evidence and accepted practice supports that estimating incidence when both prevalence and duration is known is consistent with established and tested epidemiological methodology.

It is my opinion that study-level statistically significant results are not required to develop a strong scientific opinion that the prevalence estimates from the Ganzini, Sewell and Matson research results are valid and reliable for the sub-populations researched, especially since they are consistent with finding from the CAPS and because causality was previously determined using the Bradford-Hill criteria.

It is my opinion that the results from specific sub-populations do not need to be generalizable to the general population to signal a true public health problem.

It is my opinion that the evidence supports that the Food Drug and Cosmetic Act, as amended, and all the implementing regulations have an ethically based patient safety foundation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to monitor and assure that their product is being used safely for intended uses. It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to monitor the adverse consequences associated with the use of their products. It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to comprehensively assess signals of risk raised for their products and promptly notify the FDA. It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to effectively mitigate any identified risk associated with the use, on label or off label, of their products. It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that all new drug Applicants, NDA or ANDA, have a regulatory and public health obligation to assure that their products are accurately labeled for safe use, at all times. It is my opinion, based on the material that I reviewed, that PLIVA failed to meet this obligation.

It is my opinion that there is no provision of the Hatch-Waxman amendment to the FDCA that overrides the basic core and ethics-based safety provisions (1938 and 1962) of that Act.

It is my opinion that the evidence supports that MCP-induced TD due to unapproved off-label use remains an unmitigated public health problem to this day.

It is my opinion that PLIVA has not acted as a reasonable and prudent drug manufacturer and breached their standard of care. Given clear evidence of causality, the well-documented use greater than 12 weeks, and given the strong probability of increased risk for TD conditional on that greater exposure PLIVA, should, at a minimum (1) assess the risk of TD beyond 12 weeks of exposure to determine if it is no different than that seen within the approved shorter termed exposure where the benefit-risk balance is considered by the FDA to be acceptable; (2) affirm efficacy beyond 12 weeks for the relevant indication(s) and reestablish that benefit - risk balance with the results from choice #1. PLIVA had a responsibility to petition the FDA to update the label and reflect new information. Yet, PLIVA has chosen to do neither. As a result, persons continue to be harmed to this day as demonstrated by Kenney (2008). PLIVA fell below its standard of care in not reporting to the FDA and treating physicians the data available regarding long term use of metoclopramide. If PLIVA would properly inform the FDA about the risk of TD then thousands of injuries from this drug could be prevented. Metoclopramide - as it was marketed and distributed by Defendant PLIVA at the time of Mrs. Pustejovsky's ingestion of the drug - was unreasonably dangerous and the information and warnings supplied with the drug were inadequate.”

My opinions are based on the extensive set of materials that I reviewed, however I reserve the right to add upon or update my opinions as relevant new materials from the discovery process become available to me.

_______________

Robert C. Nelson, Ph.D.