To say that Dr. Parisian doesn’t approve of Johnson & Johnson’s conduct with respect to the Ortho-Evra patch would be an understatement.
104. J&J disregarded its duty and obligations as a pharmaceutical manufacturer, ignored forty-years of oral contraceptive history and science, failed to utilize good scientific methods and appropriate evaluation of its pre-clinical safety signals for design, development, testing and marketing of ORTHO EVRA. J&J's scientists in 1996, had calculated that a 20cm2 patch should be loaded with 60 mcg of EE to deliver a nominal 20 mcg EE per day based on EE skin flux of 0.04 mcg per cm2 per hour. The EE delivery values had been based on PK characteristics of ORTHO CYCLEN OC and were aimed at obtaining a steady state range of 40-50 pg/ml (Audett, 7/23/07, p 254, I. 18-20). However, J&J's development team subsequently made a determination to increase patch EE adhesive loading to 75 mcg to increase the delivery of EE. After increasing the systemic EE delivery for the patch, J&J disregarded the safety signal of increased reporting of thrombotic events during the Phase 3 clinical trials as well as disregarded a February 20, 1998 Leiden-Site report to Dr. Shangold of unacceptable estrogenic side effects suggesting high systemic EE delivery from J&J's clinical investigator Dr. Peter Van der Ark (POE 04854048). J&J also disregarded the coagulation results of CONT-006 (APPENDIX 6 30,109,178,198, 222-225). Furthermore, during its pre-NDA negotiations with the FDA, J&J failed to adequately inform the FDA that as early as 1998 it had been aware of safety signals of increased reports of pulmonary emboli greater than anticipated for a 35 mcg OC. After obtaining the FDA's NDA approval, J&J continued to elect not to provide its own clinical experts , key opinion leaders ("KOL"), physicians and patients with adequate warnings about increased systemic EE delivery and reporting of VTE in its clinical trials in its original ORTHO EVRA label, and communications (APPENDIX 6 ¶ 283-288). J&J was aware that, unlike ORTHO EVRA being sold in the United States, J&J's EVRA patch sold outside the US was loaded with 60mcg of EE. EVRA, and also unlike ORTHO EVRA, had been able to support EE bioavailability to a 35 mcg CILEST OC (NRGEEP-POI-1025) as well as bioavailability to historical data for J&J's clinical trial lot 10607. J&J's actions regarding its follow-up of the safety of increased EE delivery of ORTHO EVRA has shown a disregard for ensuring patient safety and a disregard for its own duty to sell safe and effective products. Such actions by J&J have directly contributed to the injuries and deaths of healthy young women who elected to use ORTHO EVRA for prevention of unintended pregnancy.
I. QUALIFICATIONS
(See APPENDIX I- C.V., APPENDIX 2 -Legal Testimony History)
I. Since August 1995, I have been President and founder of MD Assist, Inc., a regulatory and medical consulting firm specializing in matters involving the United States Food and Drug Administration's regulation of products. I received my Medical Degree (M.D.) from the University of South Florida in 1978 and Board Certification in Anatomic and Clinical Pathology since 1989. I have been a general practitioner and President of Mountain Emergency Physicians. I received a Masters in Biology from the University of Central Florida. I am the author of FDA Inside and Out, a text about the FDA.
2. From 1991 to 1995, I served as a Commissioned Officer in the United States Public Health Service and achieved the rank of Lt. Commander. During the time period, I was primarily assigned to the Center for Devices and Radiological Health ("CDRH") at the Food and Drug Administration ("FDA"). Concurrently, I was also assigned clinical responsibilities at the Armed Forces Institute of Pathology ("AFIP"), Office of the Medical Examiner for the Armed Forces, Washington, D.C.
3. From 1991 to 1993, I was a FDA Medical Officer in the Office of Health Affairs (OHA), a staff office within the Center for Devices and Radiological Health (CDRH), FDA. In OHA, I provided regulatory support to both the FDA's Office of Compliance and Office of Device Evaluation. My responsibilities in OHA included health hazard and health risk assessment, Safety Alerts and physician and layperson communications, review of adverse event reports and medical literature and review of product labeling, promotions, advertising, and corporate records as to compliance with the Food, Drug and Cosmetic Act. I was responsible for the review of mandatory adverse event reports submitted by manufacturers, as well as the review of voluntary reports submitted by health care providers, patients and others. I presided over 162 health risk assessments convened to advise the FDA on overall health risk issues for the public and made recommendations to the FDA regarding the subsequent regulatory actions, which should be undertaken by the FDA, health care providers, users groups and manufacturers to help protect the public's welfare. My assignment at OHA specifically included identification of safety issues. I participated in mandatory recalls and participated in an administrative hearing as the FDA's expert witness.
4. An example of a drug safety issue that I helped identify and manage for the FDA as its medical officer occurred in 1992. The FDA had received reports in its databases from the medical literature of serious adverse events and deaths occurring in patients taking ACE Inhibitor (ACEI) drugs for regulation of blood pressure. From reviewing the AERs, both CDER's and CDRH's, the patients shared a common precipitating event such as blood exposure to certain types of membrane surfaces shortly after taking a dose of ACEI. As an example, it had been reported that patients taking ACEI drugs followed by recent exposure to surface-treated AN69 membrane during hemodialysis had an increased risk for anaphylactoid reactions. ACEI had been a class of drugs that had been first approved by the FDA for marketing in the 1980s. There had been relatively little safety information about these drugs. When a patient would take an ACEI and then had blood exposed to specific types of membrane surfaces, whether as hemodialysis or LDL apheresis, a sudden life-threatening anaphylactoid reaction could be triggered that did not respond to antihistamines such as benadryl. As the FDA's medical officer, I reviewed both drug and device adverse event reports, perform a health risk assessment as described per 21 CFR Part 7, and then made a clinical recommendation to FDA as a reasonable degree of medical certainty. Health care providers needed to be quickly informed by the FDA of the ACEI membrane association and the Agency and industry needed help to identify the etiology of the reaction and physicians needed recommendations for emergency treatment. No other FDA medical officer was assigned involvement in the work-up or handling of this drug/device safety issue. Working with an FDA epidemiologist, I helped design and issue the Agency's epidemiologic study to quickly obtain additional data for FDA and the involved drug and device industry. The epidemiology study was in the form of a questionnaire contained in the Agency's Safety Alert and designed to help capture the overall risk for the public, raise awareness of the issue, define the pharmacological mechanisms involved and trigger the appropriate label changes for both the drugs and devices involved in the issue in order to protect the public. As a result, I helped draft the current ACEI class drug warning about the risks of membrane surface exposure and anaphylactoid reaction.
5. From March 1993 to December 1993, I was a Medical Office in the Office of Device Evaluation (ODE), Division of Reproductive Abdominal, Ear, Nose and Throat, and Radiology (DRAERD), FDA, January 1994 through June 1995; I was one of two Chief Medical Officers in ODE. ODE, in contrast to OHA, is primarily responsible for pre-marketing evaluation of new product applications and clinical trials to support safety and effectiveness to begin legal marketing within the United States. In ODE, I participated in the review of proposed clinical trial, pre-marketing applications, including review of animal toxicology and biocompatibility data, in addition to the assigned responsibility of training new medical officers and scientific reviewers in application, clinical trial and labeling evaluation. I was the primary reviewing medical officer in charge of pre-marketing approval applications required for adherence to CDER's Drug Guidances, and for presentation at the FDA Advisory Panel with members from CDRH and Center for Drug Evaluation and Research (CDER). I was a primary author for FDA's guidance for Hemodialyzer Reuse labeling. I consulted as a medical officer on INDs for combination products including drugs and biologics. While in ODE, I conducted an additional 100 health risk assessments and was required to train medical officers as to methods for heath risk assessments, health hazard evaluations, annual report, adverse event and labeling review.
6. I was an initial instructor in the FDA's Staff College for training FDA reviewers in the design and evaluation of clinical data in investigational and pre-marketing applications. I had primary responsibility for review of marketing applications, labeling and was required to teach medical officers the process for evaluation and review required by the Food, Drug and Cosmetic Act for support of product marketing. I was charged with training medical officers on the process for health risk assessment and health evaluation per 21 CFR Part 7.
7. Regarding post-market surveillance of marketed products, I participated with the FDA's District Offices, Office of General Counsel, and the Office of Compliance in the review of manufacturing records, labeling, product complaints and adverse event reports obtained by the FDA. I was the primary clinician involved in several of the FDA's Major Corporate-Wide Actions for which I received various citations and honors for my services to the FDA. My awards have included Department of Health and Human Services and the Food and Drug Administration Employee of the Month.
8. I was sent by the FDA to serve as an official Agency representative to medical meetings and seminars to help identify and monitor conduct of manufacturers for potential deviations from regulations governing promotional activities. At those interactions, I was required to provide official guidance as to the FDA's interpretation of Food and Drug Laws as they pertain to all medical products and the roles of manufacturers and health care providers.
9. While at the FDA, I helped draft agency documents, guidance documents regarding requirements for obtaining FDA's marketing approval, FDA Safety Alerts, provided the FDA comm
ents for volunta
ry warnings and physician and user notifications, and agency comments on voluntary industry standards. I was a FDA liaison with the National Institutes of Health (NIH) for issues involving ENT, Renal, Respiratory, Women's Health, and Alternative Medicine. I was required to provide support to Health Care Financing Agency (HCFA now CMS) regarding the FDA's approval of product and issues involving hemodialysis. I was assigned responsibility for product adverse event reporting to the Department of Defense (DOD) and Veterans Administration.
10. One of my assigned responsibilities at the FDA, based on my clinical training and experience, was to review facts contained in product marketing applications, clinical trials, medical literature, reports of post-marketing experience, and available manufacturing documents gathered by the FDA or provided to the FDA by the manufacturer or other regulatory agencies, and then to use those facts to I) make a clinical determination to a reasonable degree of medical certainty for FDA per the FDCA and; 2) recommend the next courses of action available to FDA to protect the public health. I was also required by the FDA to advise and train other FDA employees regarding the review of facts of a case or issue, the requirements of the Food. Drug and Cosmetic Act, and making a determination to a reasonable degree of medical certainty regarding the clinical impact of the agency's actions to the public. This was a process I was trained in and required to perform for the FDA. The health risk assessment process is further described in 21 CFR Part 7. During my tenure at the FDA, I reviewed hundreds of marketing applications for safety and efficacy as well as proposed draft labeling. In this capacity, I worked with industry scientists and academic clinical investigators for evaluation, marketing and labeling review of new products. I organized national conferences with industry and physicians to discuss and obtain expert consensus regarding the development of new products and labeling as well as evaluating existing products on the market for safety and efficacy.
11. At the Armed Forces Institute of Pathology (AFIP), Office of the Medical Examiner, I was required, again based on my clinical training and experience in pathology, to take all available facts surrounding a patient's death and any involved adverse events and make a final determination: I) to a reasonable degree of medical certainty, as to the cause of death, and 2) to recommend the next steps that should be taken by the military or another agency of the federal government. In that capacity as a Medical Examiner, I provided support to the various legal staff of the armed services, as well as the FBI and CIA. While a medical examiner at the AFIP, I determined that a cause of civilian patient deaths, occurring in military hospitals, to a reasonable degree of medical certainty, appeared to have been associated with unanticipated drug/device effect. I then reported my findings as a MedWatch report to the FDA. As an AFIP Medical Examiner, I was able to trigger the FDA to investigate a major drug regulatory safety action which resulted in the protection of public health.
12. After leaving the FDA, and founding MD Assist, Inc., I have continued to provide information to individuals, manufacturers, and organizations of the FDA's requirements, Adverse Event Reporting, and labeling of FDA-regulated products. Those products have included INDs, NDAs, IDEs, PMAs, and 510(k)s for devices, biologics and drugs. I was requested by the FDA to participate in a 1997 panel of experts convened by the FDA to comment on changes proposed in the requirements for medical device labeling. I continue to consult for manufacturers, lecture at conferences and seminars regarding FDA, pre-market clearance, design of clinical trials, product labeling, Corrective and Preventive Action (CAPA) and Quality Systems.
13. I have attached a list of my last 4 years of deposition and court testimony Appendix I. A copy of my most recent Curriculum Vitae. is attached as Appendix 2. I receive $350/hr for study and $475/hr for deposition and trial testimony.
II. ROLE OF THE FDA in OVERSIGHT OF PHARMACEUTICAL PRODUCTS
(See APPENDIX 3 FDA & HUMAN PRESCRIPTION DRUG; APPENDIX 4 HISTORY of BIRTH
CONTROL PRODUCTS IN THE UNITED STATES)
14. The United States Congress mandated that the Federal Food and Drug Administration (FDA) be the federal government agency assigned responsibility for oversight of the marketing of new drugs in the United States. The authority of the FDA came directly from Congress through its passage of the Federal Food and Drug Act of 1938 ("FDCA" or "the Act") and subsequent amendments. The Act provides the FDA and the regulated-industry with a minimal standard of conduct to allow manufacturers to sell human drugs in the United States. After 1938 all new drugs were to be approved for U.S. marketing based on the FDA's prior review and approval of a marketing application containing safety and manufacturing data. The Kefauver-Harris Amendment of 1962 changed the oversight role of the FDA for human drugs. After 1962, all "new" drug applications (NDAs) were to be first reviewed by the FDA to determine if there was substantial evidence provided by the applicant for support of safety and efficacy for an intended use prior to the FDA giving approval for marketing in the United States. To address drugs which were already on the market prior to 1962, the amendment contained a "grandfather" clause and a requirement for the FDA to perform a retrospective review of 1938-1962 approved NDA drugs (based on safety only) for support of efficacy for an intended use as well as the drugs that were identical, related and similar to such NDA drugs (Drug Efficacy Study Implementation or "DESI"). DESI drugs are not to be confused with "grandfathered" drugs which are generally recognized as safe and effective (GRAS/E). Prescription "Wrap-Up" drugs are drugs that were on the market based on a claim of being either a pre-'38 or pre-'62 product or identical, related, similar to such a product.
15. The FDA's DESI review process officially began in 1966. The FDA contracted with the National Research Council-National Academy of Sciences (NRC-NAS) to help with the Agency's review of the support for efficacy and to make recommendations to the FDA. The DFSI process has included 3,443 drug products approved from 1938 and 1962 and approximately 15,000 unapproved drugs that were judged as similar to DESI drugs.
16. The Agency's current requirements for manufacturers to conduct human clinical trials prior to obtaining FDA approval providing valid scientific evidence supporting safety and efficacy, which the public associates with the FDA's current New Drug Approval process, were actually only issued by the FDA beginning in the 1980s. The Investigational New Drug (IND) regulations (21 CFR 312) were published in 1987,( 52 FR 8831), long after the initiation of commercial marketing of combination oral contraceptives such as ENOVID (1960) and ORTHO NOVUM (1962).
17. The FDA's primary role from Congress under the Federal Food Drug and Cosmetic Act ("Act") has been to ensure public safety for all products regulated by the FDA, including foods, drugs, devices, biologics, cosmetics, or veterinary products. The FDA was not mandated by Congress to develop, design and test new drugs, write drug labels, treat patients or conduct human clinical trials. There is no FDA hospital and the FDA's medical officers do not wear white coats, use stethoscopes and engage in treating patients.
18. The FDA's role is as a government agency intended to protect public safety for the products the FDA regulates. Included in that role is the FDA's ensuring of the safety and efficacy of a drug or a new indication for an approved drug before it begins to be commercially marketed in the United States to the public. The FDA's approval of a new drug application ("NDA") is not the FDA's guarantee that a new drug will be safe and effective for an intended use when used by the public. The FDA's approval is the agency's determination, based on its own review of the information provided to it by the manufacturer within a filed marketing application, that a drug and its manufacturer have fulfilled minimal requirements of the Act to allow the start of marketing in the United States. Drugs are approved for marketing when the application and data does not meet FDA reasons for denial of approval (21 CFR 314. 125-Refusal to approve an application). According to Sec. 314.105, approval of an application, (a) "The FDA will approve an application and send the applicant an approval letter if none of the reasons in 314.125 for refusing to approve the application applies."
19. Within its role of protecting the public, the FDA is also charged by Congress with monitoring reports it receives associated with both pre-approval and post-approval drug performance. The FDA receives annual reports, adverse experience reports, unpublished and published reports, reports from other agencies and governments, to help it identify drug trends and safety signals and for the Agency to take steps available to it to help protect the public. When the Agency becomes aware of a potential safety trends with a drug based on risk, the FDA is to take whatever means are currently available, feasible, practical, both mandatory and voluntary, to help protect the public. The traditional mechanisms available to the FDA, and based on agency resources and priorities and public risk, include issuance of publicity, safety alerts, warning letters, public health advisories, advisory meetings with outside experts, notification and communications with other government and regulatory authorities.
20. At the present time, the FDA has not been given direct authority from Congress to mandate or require that a pharmaceutical manufacturer conduct additional clinical studies to address safety issues that have occurred or been alleged to have occurred with an approved drug. Until the FDA Revitalization Act of 2007 and a new ability to issue fines, the FDA did not have any ability to compel phase 4 commitments, pre-approval commitments made formally between a pharmaceutical manufacturer and the FDA, be conducted, completed and reported back to the FDA. The FDA usually opts to take actions readily available to it, such as withholding or delaying approval of a new indication, threatening to issue a safety alert or negative publicity, holding a public panel meeting to discuss the safety issue, sending a Warning Letter, attempting to encourage a manufacturer to voluntarily perform additional testing or studies, updating product labeling and marketing, and/or sending out communications to health care providers or patients, in order for the manufacturer to return itself back to compliance with its own duties and obligations under the Act.
21. The FDA's employees in all of its review Centers, including the Center for Drug Evaluation and Research (CDER), must examine data provided to them by competing manufacturers with dispassion. In terms of drugs, the FDA was not designed by Congress to compete or interfere with the US pharmaceutical industry. The FDA's mandate from Congress is as a gatekeeper to protect the public health and to ensure compliance with minimal standards of the Act. After the passage of the Prescription Drug User Fee Act of 1992 (PDUFA) by Congress, in return for payment of users fees from pharmaceutical manufacturers to help with the funding of CDER, the Agency was required to re-focus and streamline Agency procedures and processes to expedite new drug approval. PDUFA changed the "customer" for the services of FDA from the public to pharmaceutical manufacturers providing new drugs to the public. All marketing applications filed with CDER have a date specified for the Agency reviewers to have finalized the Agency's action on the request. For example, in return for the user fee paid to the FDA for review and processing of an original NDA, Agency is to bring final closure to the application by a 10 month closure date.
22. For CDER to comply with its Congressional mandate, it is responsible for evaluating pharmaceutical data for new dr
ugs before they ma
y enter the United States market. The data is provided to the FDA by a manufacturer in the form of written new drug marketing applications ("NDA"). CDER must determine if there is substantial evidence to support safety and efficacy of the new drug to allow a company to begin marketing it in the United States. CDER's evaluation includes a risk versus benefit determination for the intended patient based on the provided data in the submission, including the medical literature. CDER must decide if the risks of a new drug is proportional to benefits for the indicated population.
23. For example, the FDA's CDER, Division of Reproductive and Urological Products (DRUP) was to make a final determination within 10 months, the PDUFA due data, for either approval or denial of approval of the start of commercial marketing of ORTHO EVRA in the United States. The FDA was to determine, based upon the Agency's review of the data provided to it by J&J in NDA 21-180 within 10 months, if the benefits of ORTHO EVRA transdermal patch, namely for prolonged use by healthy sexually mature women for prevention of unanticipated pregnancy, were commensurate with the risks as described to FDA in J&J's clinical trials provided in the NDA.
24. The FDA does not have a means to independently test the use of the ORTHO EVRA patch in women, but must rely on J&J to have provided it with full, truthful and accurate disclosure and discussion of the risks and benefits of ORTHO EVRA. The FDA must also rely on J&J's skill and knowledge as a major pharmaceutical manufacturer to have performed the appropriate development and product testing prior to having filed a marketing application with the FDA. The FDA also must rely on the skill and knowledge of J&J to have implemented and validated the necessary manufacturing controls to be able to produce a safe and effective commercial patch.
25. According to Sec. 314.105, approval of an application, (a) "The FDA will approve an application and send the applicant an approval letter if none of the reasons in 314.125 for refusing to approve the application applies." The FDA reviews J&J's own submitted NDA data to determine if there appear to be reasons for it to refuse to approve the ORTHO EVRA marketing application (21 CFR 314.125). The FDA's approval process has not been designed to identify fraudulent or misleading data or failure of a firm to provide data to the FDA. The FDA considered prior to approval in November 2001, whether the potential benefits of prevention of unanticipated pregnancy by a 7 day patch, the subject of J&J's clinical trials and NDA, appeared to be proportional to risks of the patch, including initial reports of VTE, stroke, death, unintended pregnancy, and dermal effects, as well as the known risks already associated with combination contraceptive products, bioavailability, pharmacokinetic and pharmacodynamic data. The FDA's reviewers also recommended that J&J's ORTHO EVRA labeling provide physicians and women warnings about the unknown risks for a transdermal combination contraceptive patch.
26. The limited and designed clinical trials, with inclusion and exclusion criteria, which are required for obtaining NDA approval are not designed to be able to capture all risks, or the frequency of risks, which will occur with wider commercial use of the product by the public. Once a new drug is marketed, the FDA monitors filed adverse experience reports (AERs) of drug performance trying to identify safety signals or trends in reporting to help take steps to safeguard the public. The FDA's AER drug database, however, is a passive surveillance database. Despite mandatory reporting requirements for the pharmaceutical industry, there is recognized under-reporting by the pharmaceutical industry with estimates of 3-10% of reports actually filed with the FDA. Since the FDA's database is a form of passive surveillance, it receives reports from all sources and accepts reports with limited amounts of patient and event data. The FDA's database is not designed to provide actual frequency of occurrence rates for events but is to be examined looking for trends in reporting for drugs and manufacturers. As an example, in 2003, the FDA. at the request of the medical officers in DRUP, began to periodically monitor AERs received for ORTHO EVRA looking for post-approval safety trends and changes in frequency and severity. FDA continued to monitor the AERs for ORTHO EVRA for post-approval trends through January 2007.
27. The FDA negotiates the contents of a drug label with each manufacturer prior to Agency approval. As discussed in the Institute of Medicine's (IOM) The Future of Drug Safety, Promoting and Protecting the Health of the Public, National Academies Press,2007. a study commissioned by the FDA, the FDA was not given authority by Congress to be able to require a pharmaceutical company to perform additional safety studies for its already marketed drug. Because of limited agency resources and authority, the FDA again attempts to compel voluntary actions from manufacturers to update its own product labeling, to voluntarily conduct post-approval studies, to warn physicians and patients about risks and to withdraw unsafe drugs from the market.
28. Regarding that responsibility, in J&J's September 29, 2000 comments (POE 00506689-720) to the FDA concerning the FDA's proposed draft for Combined Oral Contraceptive- Labeling for Healthcare Providers and Patients Docket 00D-1350, J&J's general comments began:
While we support efforts to write useful, clear product labeling that best facilitates safe and effective us of drug products, there can be disadvantages to oversimplifying language regarding risks ….. As a commercial marketer, we believe it is our responsibility to objectively relate the risks of oral contraceptive use so as to fairly and adequately inform those who prescribe our product (POE 00506694).
a. FDA AND ORTHO's ORAL CONTRACEPTIVE PRODUCTS
29. See APPENDIX 4 HISTORY of BIRTH CONTROL PRODUCTS IN THE UNITED STATES for a more detailed discussion of the history of birth control products in the United States.
30. Based on a series of public hearings beginning in the early 1960s regarding the safety of long-term prescription of combination oral contraceptives (COCs) in healthy women, to help increase patient safety, the FDA worked with the contraceptive pharmaceutical industry to help reduce the amount of laboratory and human testing data necessary for manufacturers to be able to bring new "lower hormone strength doses" of combination oral contraceptives to the market.
31. On April 4, 1984 as NDA 018-985, ORTHO NOVUM, which had been commercially available since 1962. was approved as an OC by the FDA allowing a change in active estrogen source from mestranol to ethinyl estradiol ("EE"). The ORTHO NOVUM COC would now contain ethinyl estradiol (EE) * a progestin, norethindrone. Ethinyl estradiol (See APPENDIX 4 ¶ 53-58- for further discussion of Ethinyl Estradiol) was synthesized in such a manner to reduce the liver's first pass effects and to optimize the systemic effects produced by the estrogen present in the COC. A lower delivered dose of EE in the OC could produce either an increased or comparable systemic effect by reducing the metabolism by the liver.
32. Wyeth obtained approval of a "triphasic" OC TRIPHASIL (NDA 019-192) on November 1, 1984 which contained EE (0.03mg,0.04mg,0.03mg)* progestin levonorgestrel (0.05mg,0.07mg, 0.125mg) [*Note: Wyeth's TRIPHASIL would be used as a comparator in J&.J's Phase 1 and 3 clinical studies- NRGEEP-PHI-CONT-004, NRGEEP-PHI-017for approval of the ORTHO EVRA NDA (See APPENDIX 5) ].
33. J&J introduced a new OC which it called "ORTHO-CYCLEN" that used J&J's new progestin "norgestimate" (NGM). ORTHO-CYCLEN [ EE*NGM] was approved by the FDA as NDA 19-653 on December 29, 1989. [*Note: The ORTHO-CYCLEN NDA and data is referenced (bridged) by J&J for obtaining approval of ORTHO EVRA's NDA]. J&J also entered into the "triphasic" market using norgestimate. J&J obtained the FDA's approval of NDA 19-697, July 3, 1992, for marketing of ORTHO-TRICYCLEN. [*Note: The ORTHO-TRI-CYCLEN NDA and data is referenced (bridged) by J&J for obtaining approval of ORTHO EVRA's NDA].
34. On November 14, 1995, the FDA issued FDA Talk Paper T95-61- Oral Contraceptives and Risk of Blood Clots written by Susan Cruzan. The FDA was responding to inquiries about the risks of rare and nonfatal blood clots (or venous thrombotic events ("VTE") reported from use of combination oral contraceptives (COCs) containing 3rd generation progestins, desogestrel or gestodene. The FDA had concluded from its internal review of three recent unpublished studies that the risk was not great enough to justify switching a woman to other products. The progestin gestodene had not been approved for products marketed in the United States. The progestin desogestrel was marketed in the United States as DESOGEN (Organon, Inc.) (*Duramed's Mircette/Mercilon) and ORTHO-CEPT (Ortho-McNeil Pharmaceutical). In 1995, according to the FDA, these two OC products currently represented about 15 percent of the American oral contraceptive market (APPENDIX 4 ¶30-50).
35. The FDA intended to work with manufacturers to update the risk information in the OC product labeling. The relevance of this article to discussion of ORTHO EVRA is twofold. First, Dr. Shangold, an OB/GYN and J&J's Clinical lead for ORTHO EVRA development, during his 2007 deposition referenced J&J's 1995 concerns about VTE raised with introduction of 3rd generation progestin desogestrel, ORTHO-CEPT. He indicated it was prudent for J&J to have developed a program of vigilance for VTEs (Shangold, June 5, 2007 p148 1.14-24). Also, the FDA provided the average annual risk rate of non-fatal venous thromboembolism (VTE) for 1995:
–4 cases per 100,000 in healthy women who are not pregnant and not taking hormones.
-10-15 per 100,000 for women taking older, low dose oral contraceptives
–20-30 per 100,000 for women taking desogestrel and gestodene containing products
–60 per 100,000 women for pregnant women.
36. (See APPENDIX 3 FOOD AND DRUG ADMINISTRATION (" FDA") & Human Prescription Drugs; See APPENDIX 4 – HISTORY OF BIRTH CONTROL PRODUCTS IN THE UNITED STATES)
b. ROLE of the HUMAN PRESCRIPTION DRUG MANUFACTURER (See APPENDIX 3)
37. The role of the FDA with the drug process has already been discussed and it is primarily involved with a new drug to respond to a manufacturer's request for its approval to use the product for treatment of the public in clinical trials or commercially. All sponsors of FDA-regulated products, whether it is a food, drug, biologic, cosmetic, device, or veterinary product, are required to ensure that the products it sells to the public in the United States remain safe and adequately and truthfully labeled and promoted. Sponsors of all FDA-regulated products have the duty and obligation to continue to monitor the safety of products sold to the public and must have in place functional mechanisms to respond to safety issues. FDA-regulated products are to provide adequate and truthful labeling and warnings, as well as be advertized and promoted with fair and well-balanced information.
38. In the case of human pharmaceuticals, it remains the drug manufacturer's responsibility not the FDA's, to ensure throughout the product's lifecycle compliance with applicable sections of the Act. It is the drug manufacturer's responsibility and duty, not the FDA's, to have performed adequate drug design, testing, development and manufacturing of a commercial product as well as post-approval monitoring, surveillance and voluntary updating of warnings.
39. As a responsible manufacturer would be aware, the FDA does not design, test and develop new drugs. J&J's drug development process for ORTHO EVRA was independent of the FDA. ORTHO EVRA and J&J's actions did not rest on the FDA's having requested all the types of testing that needed to be done to develop ORTHO EVRA. Design and development of a new product like ORTHO EVRA, as with any other pharmaceutical manufacturer and a new drug, is within the purview of the manufacturer and not the FDA. Each manufacturer, not FDA, is responsible for its own actions and conduct during drug development, manufacturing and promotion under the Act.
40. When a drug sponsor desires to improve the safety and effectiveness of its drug, or to improve its label and marketing to ensure its own compliance with the Act or to protect public safety, the Act provides the drug manufacturer an ability to voluntarily add or strengthen its warnings or otherwise change its label or marketing without receiving the FDA's prior approval. The manufacturer can also voluntarily send out timely communications, such as Dear Doctor Letters, to health care providers and patients, as well as train its own sales representatives to help ensure public safety. Neither the Act nor the FDA would require a manufacturer to sell drug products that are unsafe and ineffective and pose an increased risk to the public.
41. A drug sponsor, not the FDA, is responsible for its own product's performance, pre-and post-approval, as well as adequacy and truthfulness of its communications with the FDA, labeling and promotions. Each drug sponsor is to have in place fully functional pharmacoviglance methods for its own monitoring of reports and for reacting quickly to post-approval performance issues for a drug through its complete lifecycle. Pharmacovigilance is generally regarded as all post-approval scientific and data gathering activities relating to the detection, assessment, understanding, and prevention of adverse events or any other product-related problems. Pharmacovigilance includes the use of pharmacoepidemiologic studies which include methods such as standard case-control, observational or epidemiological studies, to help it voluntarily update drug labels, warnings and communications to help ensure a drug remains safe and effective. Obtaining the FDA's approval to market a drug is a small component in the scheme of the drug's complete lifecycle, or the tip of the iceberg. In contrast to the limited role of FDA for approval and monitoring, the manufacturer under the Act has the larger role to remain actively engaged in every aspect of its drug's safety and efficacy throughout the drug's lifecycle.
42. Good pharmacovigilance practice (GPP) begins with a manufacturer is the ability to acquire complete data from spontaneous adverse event reports (AERs) it receives. The process includes implementation of good case report forms and medication error reports. Effective case series development and assessment by the manufacturer looks for causality of safety signals. A safety signal can become apparent with an excess of adverse events being received from various sources associated with a product's use or it can be a single well-documented case report, preclinical findings, or experiences derived from similar types of products. Safety signals may be further assessed by the manufacturer in terms of magnitude, population at risk, changes in risk over time, biological plausibility and other factors. A process called data mining can be used as a signaling tool.
43. Pharmacovigilance data mining uses raw case report data and arrays of all drug-adverse event combinations. The manufacturer is able to calculate the "expected" (E) number of events for all drugs, then can compare each drug's "observed"(O) event to "expected". Various methods may be used to generate relative signal scores ("intensity") of ["O:E"]. Data mining has been a useful adjunct tool to routinely evaluate safety signaling methods. The results of data mining can trigger further exploration of trends and validation of the signal by a manufacturer.
44. Safety signals can be new unlabeled adverse events, an observed increase in severity or specificity of a labeled event, an observed increase in the frequency of a labeled event, new interactions or events associated with confusion of a product's name, packaging, or use. Safety signals may be associated with demographics- age, gender, race. Safety signals can also be associated with the duration of exposure and dose. Safety signals can have a relationship between concomitant medications and potential interactions and the risk of event as well as a relationship between co-morbid conditions and the risk of the event. Safety signals can show effects as lot-to-lot variation or in differences in product formulation and risks for the event. Safety signals can utilize estimates of the potential magnitude of risk or differences from known background rates.
45. A responsible pharmaceutical manufacturer will promote its drug's use to health care providers and patients by fair and balanced marketing and will not disregard the requirements of the Act. A responsible drug sponsor does not promote its drug for new off-label/unapproved uses, particularly when there is data that suggests the use is neither safe nor effective. A responsible manufacturer would not provide physicians and patients with false and misleading labeling, promotions, advertising and communications whether through its sales representatives, or through use of outside public relations agencies, 3rd parties, and paid advocates and key opinion leaders. A responsible manufacturer would not delay in providing FDA, physicians and patients with risk and lack of efficacy information.
c. PURPOSE of the NDA PROCESS
46. The FDA is mandated to consider the "risk/benefit" of a drug for approval or denial of a new drug marketing application (NDA). The FDA's 21 CFR 314.2 Purpose of the NDA Process: "The purpose of this part is to establish an efficient and thorough drug review process in order to (a) facilitate the approval of drugs shown to be safe and effective; and (b) ensure the disapproval of drugs not shown to be safe and effective."
47. As the pharmaceutical industry would be aware, the FDA can "approve" an NDA (21 CFR 314.105) if none of the reasons for refusal in 21 CFR 314.125 (Refusal to approve an application) are applicable. The FDA may refuse approval of an NDA for any of the following reasons:
The first reason for FDA's NDA refusal, the methods to be used in, and … controls used for, the manufacture, processing…of the drug substance or the drug product are inadequate to preserve it identity, strength, quality, purity, stability, and bioavailability. The second reason, the investigations required under section 505(b)…do not include adequate tests by methods reasonably applicable to show whether or not the drug is safe for use under the conditions prescribed, recommended, or suggested in its proposed labeling. The third reason for NDA refusal, the results of the tests show that the drug is unsafe for use under the conditions prescribed, recommended, or suggested in its proposed labeling or the results do not show that the drug product is safe for use under those conditions. The fourth reason, there is insufficient information about the drug for the FDA to determine whether the product is safe for use under the conditions prescribed. The fifth reason, there is lack of "substantial evidence" consisting of adequate and well-controlled investigations (per Section 21CFR 314.26) that the drug product will have the effect it purports or is represented to have under the conditions of use prescribed. recommended, or suggested in the proposed labeling. The sixth reason for FDA's refusal of N DA approval is that the proposed labeling is false or misleading in any particular. The seventh reason, the application contains an untrue statement of a material fact.
The eighth reason for the FDA's denial of approval is that the drug product's proposed labeling does not comply with requirements of labels and labeling in Part 201. The ninth reason for FDA refusal, the application does not contain bioavailability or bioequivalence data required under Part 320.
Skipping to the 13th reason for the FDA's refusal to approve an NDA application, the methods used in the manufacture, processing…of the drug substance or drug product do not comply with the current good manufacturing practice ("cGMP") regulations in Parts 210 and 211. The fourteenth reason for refusal, the application does not contain an explanation of the omission of a report of any investigation of the drug product sponsored by the applicant, or an explanation of the omission of other information about the drug pertinent to an evaluation of the application that is received or otherwise obtained by the applicant from any source.
48. Both before and after a manufacturer obtains its NDA approval for a drug, the FDA allows a manufacturer the ability to voluntarily change the label pursuant to a "Changes Being Effected (CBE) Supplement", to change a label that "deletes false or misleading information, or unsupported indications for use or claims of effectiveness (21 CFR 314.70 ( c)(2)(iv). Therefore, at any time after NDA approval, J&J, in terms of ORTHO EVRA, could have behaved as a responsible pharmaceutical manufacturer and voluntarily ensured its own product's compliance with the Food, Drug and Cosmetic Act and protected public safety.
III. OPINIONS:
49. The following opinions are all rendered to a reasonable degree of medical, scientific and regulatory certainty, based on my training, expertise, and education and following my review of relevant documentation. I reserve the right to amend my opinions subsequent to additional discovery. (See also APPENDICES 3. 4, 5 and 6 as additional BASES FOR MY OPINIONS.)
IV. BASES for OPINIONS- Please see the following documents:
A. APPENDIX 3- FOOD & DRUG ADMINSITRATION ("FDA") & HUMAN PRESCRIPTION DRUGS
B. APPENDIX 4- HISTORY OF BIRTH CONTROL IN THE UNITED STATES
C. APPENDIX 5- J&J's ORTHO EVRA's CLINICAL STUDIES
D. APPENDIX 6- HISTORY OF ORTHO EVRA PATCH
OPINION #1:
50. Johnson & Johnson ("J&J") delayed in fulfilling its duties and obligations as a responsible United States pharmaceutical manufacturer when it did not voluntarily ensure its own timely compliance with the requirements the Food and Drug & Cosmetic Act ("Act"). J&J marketed ORTHO EVRA with false and misleading information to the FDA, physicians and women without adequately warning that ORTHO EVRA delivered increased serum ethinyl estrogen ("EE") when compared to a 35mcg oral contraceptive. J&J also failed in its obligations and duties when it did not adequately warn women about increased risks of unintended pregnancy, venous thrombotic events (VTE) and death associated with ORTHO EVRA. J&J's actions, inactions, and disregard for public safety directly contributed to injuries and deaths of healthy young women who electively opted to try ORTHO EVRA as a method of birth control. J&J failed and/or delayed in providing FDA, physicians and/or women with the following warnings and product information:
#1: J&J's ORTHO EVRA transdermal patch does not deliver "20 mcg ethinyl estradiol (EE) per day to a woman"(APPENDIX 6 ¶445)
#2: Use of ORTHO EVRA transdermal patch is not comparable to using a 20 mcg EE oral contraceptive. At a June 29, 2004 meeting held between J&J, the FDA's medical officers requested that J&J revise its ORTHO EVRA label to clarify the EE exposure to women, the FDA indicated to J&J that the label reference to 20mcg was "misleading" in terms of EE exposure and an oral contraceptive. (POE 001027767-71) (APPENDIX 6 ¶350). On July 29, 2004, the FDA wrote a follow-up letter to J&J requesting that they revise the label about EE delivery ( APPENDIX 6 ¶ 351).
#3: ORTHO EVRA's 1999 Phase 1 bioavailability study, NRGEEP-PHI-014 incorrectly applied a 60% correction factor, reducing measured systemic EE by 40% (APPENDIX 6 ¶130-142).
#4: A pre-commercial launch article, Fertility and Sterility, Supplement 2, February 2002 Vol. 77 No.2 Pharmacokinetic Overview of ORTHO EVRA/EVRA by J&J's L.S. Abrams, Ph.D, Donna Skee, M.S, Jaya Natarajan, Ph.D, and Frankie Wong, pp S3-S 12, did not indicate that in 1999 that J&J had used a 60% correction factor to calculate serum EE delivery, reducing measured EE values by 40%. (Exhibit 681). (APPENDIX 6 ¶130-142).
#5: ORTHO EVRA delivers 30.4mcg of ethinyl estradiol (EE)( abdomen) and 38.1 mcg of EE (buttock) which would be equivalent to a women taking oral contraceptives with 60.8mcg (abdomen) or 76.2 mcg (buttock) of EE. In contrast, the bioavailability of ORTHO TRI-CYCLEN delivers 17.5mcg of EE to a woman's systemic circulation (POE 08372944). (APPENDIX 6 ¶130-142).
#6: ORTHO EVRA can deliver systemic EE to women greater than if they were taking 50 mcg EE oral contraceptives(APPENDIX 6 ¶443,469,474). Since 1988, all oral contraceptives with greater than 50 mcg of EE (delivering greater than 25 mcg EE systemically) have been withdrawn from the market due to unacceptable risk with no additional benefit (APPENDIX 4).
#7: The FDA appears to remain inadequately informed by J&J of the inappropriate use of a 60% correction factor in its 1999 bioavailability study. J&J's "Problem Statement", Dr. Heald's May 3, 2005 memo, makes no reference to J&J's duty or obligation to inform the FDA, any other Health Authority or physicians that its calculations were wrong and created misleading information regarding estrogen delivery from the patch (POE 04818681-90)(APPENDIX 6 1130- 142). J&J testimonies in 2007 continue to support that the FDA, other Health agencies and physicians have not been adequately informed by J&J of the incorrect EE calculations. An August 2006 J&J internal email from James Oldham to Alfred Tonelli was regarding his providing an update for ORTH EVRA delivered EE dose "Since clinical/clinical PK cannot say for sure what is delivered', the FDA asked them to take references to an amount from the label (POE 09107992-3) (APPENDIX 6 ¶ 529). In June 2006, Health Canada's reviewer questioned J&J's method used for calculating estrogen delivery in NRGEEP-PHI-014 (POE 07597415- 6, Exh 986) (APPENDIX 6 1514). After receiving J&J's response, Health Canada required J&J to perform a new mass balance phase 1 study to determine EE delivery for EVRA being marketed in Canada (NRGEEP-CONT-1012)(POE 07573541- 605).
#8: The FDA first requested during the August 18, 2005 label negotiations with J&J that J&J update ORTHO EVRA's label to warn physicians and patients that systemic EE delivery and estrogenic effects may approximate systemic EE delivery of a 50mcg oral contraceptive(POE 02005038-9; POE 04401871-3) (Exh 232); (POE 05059292-5) (APPENDIX 6 ¶ 443,469,474). J&J elected not to accept the FDA's recommendation for its label. (POE.SP.000107–120)(Exh P-153) (APPENDIX 6¶480-483).
#9: J&J's 1999 Phase I study, NRGEEP-PHI-017, a pharmacokinetic comparison of ORTHO EVRA to three different oral contraceptives, measured systemic EE delivery for ORTHO EVRA as twice as high as Triphasil and three times as high as Alesse or Mercilon. Despite the urging of EMEA's Expert Reviewer, Dr. John Guillebaud upon his first learning of the results in 2000 after recommendation of approval of EVRA, he urged transparency and making the complete results public (POE 01039910-11). Both of these steps were not yet taken by J&J from documentation reviewed. (APPENDIX 6130,143,167-70, 179-188,367-369). When the complete results of PHI-017 were requested by New Zealand's Health Authority as a result of J&J's request to market EVRA in 2004, the EVRA application was subsequently denied December 2004 based on the additional data from NRGEEP-PHI-017((POE 02271757)(APPENDIX 611367-369).
#10: In 1998 during ORTHO EVRA clinical trials, J&J received two reports of women with pulmonary emboli, producing a VTE risk of 118/100,000 Woman Years, an increase from the VTE risk for 35 mcg oral contraceptives. The FDA was not adequately informed about the two PEs during pre-NDA negotiations, July 1999. The FDA was notified of a "safety concern" of thrombotic events by telephone and letter September 3,1999 with a follow-up letter 6-months later, March 2000, informing FDA of two PEs in the clinical trial. (APPENDIX 61 144-149, 151-164)
#11: The risk of venous thrombotic events (VTE) in the ORTHO EVRA clinical trials was 118/100,000, increased from the VTE risk for 35 mcg oral contraceptives(APPENDIX 6 1 23, 257).
#12: The risk of VTE in the ORTHO EVRA clinical trials is not 59/100,000WY, the risk given by J&J to clinical investigators in Investigator Brochures and to J&J's Global Experts Panel members and Key Opinion Leaders at a February 8, 2004 meeting (POE 00127649)(APPENDIX 61178, 283-288).
#13: In J&J's Coagulation Study, NRGEEP-CONT-006, the change in percent prothrombin fragment F1*2 values, a marker of hypercoagulable condition, for ORTHO EVRA versus two oral contraceptive comparators, Mercilon and Triphasil, was statistically significant. In the NDA data, as indicated by Dr. Andrew Friedman during his deposition given in 2007, J&J provided FDA's reviewers only the absolute values which did not support statistical significance. Dr. Friedman indicated J&J had been aware that the percent change was statistically significant but did not provide that information to the FDA (Friedman, June, 2007, pages 363-369). Measurement of circulating levels of the prothrombin fragment 1*2 (F 1*2), according to the American Heart Association is considered a specific marker of thrombin generation in vivo, a diagnostic tool in the evaluation of the hypercoagulable state.[J.A.Paramo, et al. Prothrombin Fragment 1*2 Is associated with Carotid Intima-Media Thickness in Subjects Free of Clinical Cardiovascular Disease, Journal of American Heart Association, 2004;35; 1085-108
9](APPENDIX 6 ¶30,109
,178,198,222-225)
#14: PRI/EDN-NED-1, a post-approval study required by the EMEA, had results similar to J&J's 1999 NRGEEP-PHI-017. It showed a 60% increase in serum EE delivery by ORTHO EVRA when compared to a 35 mgc oral contraceptive. The NED-I study had used patches obtained from J&J's US manufacturing Process Validation lot. J&J's PRI/EDN-NED-1's report was finalized October 28, 2003. It was first submitted to FDA in J&J's NDA Annual Report, January 19, 2004, as "Clinical Data". J&J resubmitted the PRI/EDN-NED- I study in its May 19, 2004 Briefing Book in preparation for a meeting with FDA June 29, 2004. Dr. Heald, a former FDA employee, testified that Annual Reports are considered as routine submissions by FDA. Because of that, and J&J's upcoming June 29, 2004 meeting, he stated "that's why we decided that we better contact them with the meeting request just so NED- didn't get lost in the shuffle"(Heald 2/1/07, p.92, 11- 11) (APPENDIX 6 ¶ 322). FDA's medical officers at the June 2004 meeting, Drs Davis and Monroe, not members of J&J, were the individuals calling attention to the significance of the NED-1 data in terms of 60% increased EE delivery and patient safety (POE 0600858)(APPENDIX 6 ¶438-440, 523). FDA's Dr. Ortiz conducted the Agency's first formal pharmacokinetics review of PRI/EDN-NED-1 study data on August 2005 (FDACDER 000795-03)(APPENDIX 6¶ 438-440).
#15: PRI/EDN-NED-I showed a 60% increase in delivery of serum EE compared to a 35mcg oral contraceptive. The PRI/EDN-NED-I results were not included in ORTHO EVRA's product label until November2005 (APPENDIX 6 ¶438-440,523).
#16: J&J's November 6, 2003 Post-Marketing Safety Update had a 22 times increase in reporting of unintended pregnancy for ORTHO EVRA vs. ORTHO TRI-CYCLEN and 10 times increase in reporting cerebrovascular accidents (CVA) for ORTHO EVRA vs ORTHO TRI-CYCLEN OCs (POE 00216399-412)(Exhibit P-61)(APPENDIX 6 ¶ 270-273).
#17: J&J in June 2003 began to consider pursuing the development of a lower estrogen patch based on "safety" (POE 0807254-5)(Exhibit 676)(APPENDIX 6 ¶262).
#18: Placement of a patch at any of four recommended anatomic sites are not therapeutically equivalent. Placement of the patch on the abdomen produces a 25% reduction in systemic EE delivery. J&J did not consider abdomen placement therapeutically equivalent to the other three sites. (Abrams, June 7, 2007, p.108, 1.20-22)(APPENDIX 6 ¶ 204).
#19: Dr. Patrick Caubel of J&J, the physician most responsible for ORTHO EVRA's commercial Benefit/Risk Management, resigned from J&J on October 3 , 2005. He was aware of twenty fatalities form ORTHO EVRA when he resigned. He stated in his resignation letter addressed to Mr. Weldon, COB, J&J, that: "…the compelling evidence that the estrogenic exposure was unusually high compared to the oral product, and that the reporting rate of fatalities and over various thromboembolic disorders was out of range for this class of products." (Exhibit 179)(APPENDIX 6 ¶465-466)(*underlining and italics added for emphasis.)
#20: J&J's Form 3568 to the FDA dated June 21, 2007 still does not correctly indicate that J&J is unaware of the strength of EE delivered by ORTHO EVRA. The form incorrectly indicates the strength of EE of the ORTHO EVRA patch is 20 mcg, the value J&J management has known since 2005 was incorrectly derived from PHI-014 by use of an inappropriate 60% correction factor. (POE.SP.096014)(APPENDIX 6 ¶ 543-545).
#21: J&J's Senior management has testified in 2007 that they are not aware of the amount of serum estrogen actually delivered to a woman per day by an ORTHO EVRA patch. (Benze 3/21/07, p 199; Heald 2/1/07, p. 160-161; Soons 9/2/07, p.68-70). An August 2006 J&J internal email from James Oldham to Alfred Tonelli was regarding his providing an update for ORTH EVRA delivered EE dose "Since clinical/clinical PK cannot say for sure what is delivered", the FDA asked them to take references to an amount from the label (POE 09107992-3)(APPENDIX 6 ¶ 445, 518, 529).
#22: J&J's June 21, 2007 summary letter to FDA, submitted with a post-approval epidemiology study required by the EMEA comparing ORTHO EVRA to 30mcg EE * levonorgestrel OC, fails to accurately disclose important safety facts to the FDA about the risks of ORTHO EVRA in the study. Information and findings not fully discussed by J&J in its letter to the FDA include (POE.SP.096014-80) ) (APPENDIX 6 ¶ 543-545):
a. The age adjusted incidence rates for idiopathic VTE in current users of ORTHO EVRA were higher for ORTHO EVRA when compared to the 30ug *levonorgestrel OC (66/100,000 PY and 50.2/100,000 PY) (POE.SP.096033).
b. The age adjusted incidence rates for idiopathic VTE in current users of ORTHO EVRA was 4.5 times higher for women 40-44 years of age for ORTHO EVRA than 30ug OC (POE.SP.096033).
c. The odds ratio (OR) for ORTHO EVRA idiopathic VTE was 2.0 ( 95% CI 0.9-4.1 p___.07) and the incidence ratio (IRR) adjusting for age was 1.5(95% CI 0.8-2.7 P*0.2) when compared to 30mcg EE * levonorgestrel OCs (POE.SP.096037).
d. IRRs for ORTHO EVRA users increased substantially with increasing age while IRR for women 40-44 years using levonorgestrel OCs with 30ug EE fell (POE.SP.096037).
e. According to the authors, the OR(2.0) for idiopathic VTE comparing Ortho EVRA to levonorgestrel OC was higher than would have been expected given the results of the ORTHO EVRA versus norgestimate containing OC with 35ug had similar risks for VTE (OR 1.0) (POE.SP.096038).
f. Unlike what the FDA was told in the J&J summary, the authors indicated that, due to the lack of reports of stroke and myocardial infarction (MI), they had not been able to perform a formal study of risk. (POE.SP.096042).
# 23: The FDA's March 1, 2006 Safety Update for Adverse Event Reports (AERs) filed with the FDA for ORTHO EVRA compared to other combined hormonal contraceptives was the only product for which the FDA had received reports of cerebrovascular event (CVE) deaths with 80% of reported CVE deaths occurring in women younger than 25 years of age. ORTHO EVRA when compared to other combined hormonal contraceptives, had AERs reporting cerebrovascular events (CVEs) and myocardial infarction in younger women not identified as being overweight. The differences in AER reporting rates were most noticeable in women younger than 30 years of age (FDACDER 001193). The FDA's database is a passive surveillance database, used by the FDA to look for safety trends and cannot provide the actual rate of occurrence of events due to under-reporting by the industry and health care providers. (APPENDIX 6 ¶511-522)
ADDITIONAL SUPPORT FOR OPINION #1
a. J&J's FAILURE/DELAY TO REVEAL MATERIAL FACTS
51. Under Sec. 1.21 Failure to Reveal Material Facts, labeling of a drug shall be deemed misleading, if 1) it fails to reveal facts that are material in light of other representations made or suggested by statement, word, design, device or any combination of; or 2) material with respect to consequences which may result from use of the article under the conditions prescribed in the labeling or such conditions of use as are customary of usual. ORTHO EVRA's labeling has failed to reveal material facts to physicians and women and is misleading in that it incorrectly indicated to the FDA, physicians, and women that the ORTHO EVRA patch delivered 20mcg EE per day.
52. ORTHO EVRA's product insert is to include both a prescription insert intended for a health care provider (HCP) as well as patient insert intended for consideration by a woman. Estrogen-containing drugs are considered by the FDA to be lifestyle drugs with each woman capable of electing whether or not to use the drug based on her consideration of the recommendations from her health care provider as well as her own consideration of the risks and benefits. Per 21 CRF§ 310.515-patient package inserts for estrogens, the FDA concluded that safe and effective use of drug products containing estrogens require that each patient be fully informed by the product insert of the benefits and risks involved in the use of these drugs. Each estrogen drug product, including estrogens in fixed combination with other drugs, is to be dispensed to patients with a package insert containing information concerning the drugs' benefits and risks. An estrogen drug product that does not comply with the requirements for an adequate patient insert is considered to be misbranded under section 502(a) of the Act. The FDA issued an informal labeling guidance under 21 CFR §10.90(b)(9) to provide industry with assistance in helping meet the requirements for patient package inserts. The exemption to 21 CRF§ 310.515 includes "estrogen-progestogen oral contraceptives" which are addressed under § 310.501". Labeling requirements for estrogen-progestogen oral contraceptives as set forth in § 310.501. also require that each patient be fully informed of the benefits and risks involved in the use of oral contraceptives in the patient package insert. Any estrogen-progestogen oral contraceptive that does not comply with the requirement, as with other estrogen drugs, is considered misbranded under section 502(a) of the Act. J&J did not provide an adequate discussion of the risks for ORTHO EVRA for either the woman or her HCP in its product information ( Label, labeling and misbranding are defined in the Act as 201(k),(m)and (n)). J&J also utilized an aggressive false and misleading direct-to-consumer (DTC) and health care provider advertising and marketing campaign for its commercial launch of ORTHO EVRA but without fulfilling its duty to provide adequate risk versus benefit information.
53. There was not adequate and timely post-approval information provided to women or physicians as a result of J&J's November 2003 internal review of AERs. Neither group was adequately informed of the 22 times increased reporting of unintended pregnancy (lack of efficacy) and 10 times increased reporting of cerebrovascular accident (CVA) (increased risk) of ORTHO EVRA when compared to ORTHO-TRICYCLEN. That post-approval safety and efficacy information would have been critical to a woman's decision to use ORTHO EVRA patch as her primary method for prevention of unintended pregnancy, particularly when there were safer alternative birth control options available to her.
54. The FDA approved marketing in the United States of the ORTHO EVRA product that was described to it by J&J's NDA 21-180 in November 2001 (See APPENDIX 6). FDA's approval was based on J&J's own representations to the Agency that J&J had independently determined through its own testing and validation that its transdermal combination contraception product reliably delivered "150mcg 17d- acetylnorgestimate and 20 mcg ethinyl estradiol per day" to a healthy pre-menopausal fertile woman via a 7-day patch when applied directly to her skin for prevention of unintended pregnancy. J&J provided substantial evidence to the FDA in its NDA supporting safety and efficacy through its use of Phase 1, 2 and 3 clinical data. As a result. ORTHO EVRA was the first and only combination hormonal contraception patch approved by the FDA.
b. J&J's SENIOR MANAGEMENT REPORTEDLY IS "UNAWARE" OF THE AMOUNT OF "EE" DELIVERED TO A WOMAN BY THE ORTHO EVRA PATCH IT CONTINUES TO MARKET IN THE USA
55. By J&J's own senior management testimonies of 2007, more than six years after ORTHO EVRA's commercial market launch and after millions of ORTHO EVRA patches have been sold to women in the U.S., J&J acknowledges they are not aware of how much EE is delivered per day to a woman from an ORTHO EVRA patch. ( Benze 3/21/07, p 199; Heald 2/1/07, p. 160-161; Soons 9/2/07, p.68-70). J&J's management took steps to delay informing the FDA. physicians and patients that the risks from EE delivery by ORTHO EVRA were greater than the risks of EE delivered from a 35mcg COC. J&J delayed in informing the FDA that as early as 1998 it was aware that there was an increased risk of women developing pulmonary emboli during the clinical trials. From the documents reviewed, J&J's management was clearly aware prior to NDA approval that ORTHO EVRA delivered EE at least comparable to systemic EE delivery of a 60.8mcg (abdomen) or 76.2 mcg (buttock) oral contraceptive ("OC") with an increased risk of venous thrombotic events (VTE).
56. In contrast to the testimony of J&J senior management in 2007, J&J's FDA Form 3568 submitted on June 21, 2007 continues to indicate that the ORTHO EVRA patch delivers 20 mcg of ethinyl estradiol (POE.SP.096014), a number incorrectly derived in NRGEEP-PHI-014.
c. J&J's DELAY in TIMELY UPDATING ORTHO EVRA's LABEL
57. On May 4, 2005, there was an internal J&J discussion of a teleconference interaction with the FDA regarding its pending labeling supplement submitted on November 3, 2004. The FDA had made it clear to J&J, as supported by the PRI/EDN-NED-I data of increased EE delivery to a woman from ORTHO EVRA, that the Agency wanted the ORTHO EVRA product insert to clearly reflect to physicians that the EE delivery was not comparable to a 20mcg OC (POE 00118984- 5). The FDA was willing to approve J&J's current proposed label supplement, with the condition that a statement comparing ORTHO EVRA to a 35-mcg OC be included on the pouch label as well as in the inserts and folding carton. Dr. Benze, J&J, explained to the FDA that the Company was now evaluating all relevant PK data and would like the Agency to disregard the current proposed label supplement. He continued that J&J would rather submit a new proposed label within 2 weeks. Dr. Griebel of the FDA explained that it was important to get this done quickly., "as the PDUFA date was approaching and the Agency is under scrutiny to quickly turn around labeling supplements especially where patient safety may he affected". Dr. Fitchet, of J&J, suggested the Company provide an alternative sentence to replace the FDA's "equivalent to a 35 mcg OC". Dr. Monroe felt "this was an acceptable alternative providing it indicated that the level of EE is expected to be greater than the one achieved with a 20 mcg OC"
58. The results of the EMEA's required study, PRI/EDN-NED-I, were eventually evaluated the FDA's Steven Ortiz, R.Ph, Ph.D (FDACDER 000795-03) in August 2005. The study had been originally submitted to the FDA in J&J's Annual Report (January 2004). The NED-I summary results were eventually identified by the FDA to be deemed as important and conveyed to the Clinical Division on August 15, 2005. NED-I showed the FDA that the serum EE levels were increased by approximately 66% higher than 35mcg CILEST OC at one week and approximately 56% higher after 2 cycles.
59. By August 29, 2005, there was a General Correspondence submission sent to Dr. Shames, FDA, containing J&J's current proposed labeling changes for Agency review with a reference copy of PRI/EDN-NED-I dated August 4, 2005. This submission had been sent to the FDA as a follow-up to an earlier August 18, 2005 teleconference held between J&J and the FDA to discuss proposed revisions of the product insert (POE 04886678-81). In its August submission, J&J indicated it now intended to remove from its ORTHO EVRA insert the following statements:
…releases 150 mcg of norelgestromin and 20 micrograms of EE to the bloodstream per 24 hours. This level of transdermal release of EE results in exposure of EE greater than that produced by an oral contraceptive product containing 20 micrograms of EE.
60. Based on J&J's own internal documents and prior to NDA approval in 2001, J&J was aware, and eventually the FDA would become aware, that these label statements were false and misleading regarding ORTHO EVRA's serum EE delivery to women.
61. There was an FDA and J&J teleconference on October 19, 2005 (POE 03640436- 39)(Exh 233) with minutes recorded by J&J and circulated on November 14, 2005. At the meeting the proposed labeling changes submitted on August 29, 2005 were again discussed. Dr. Daniel Shames had underscored the Agency's commitment to have the issue resolved by October 28, 2005 but reiterated that, absent agreement, the Agency would issue a Safety Alert on the FDA MedWatch line. The Agency's counterproposal moved the PK information from Precautions to Warnings. The Agency believed that the PK information was very important and needed to be prominent in the label. Dr. Shames again reiterated the need for closure of the labeling issue by October 28, 2005. The FDA requested that J&J submit its counterproposal label so that the FDA could take final action by October 28, 2005.
62. Dr. Monroe explained that, based on the data and exposure levels and AUC compared to 35mcg pill, the FDA wished to increase the prominence of the "60% increase" in EE in the label. Based on the increased exposure to estrogen by the patch, the FDA believed this level of prominence was warranted.
63. In minutes from an October 24, 2005 teleconference held between the FDA and J&J, Dr. Shames underscored once again the FDA's commitment to resolution of the label changes by October 28, 2005 or, absent an agreement with the Sponsor, the Agency was prepared to post a Safety Alert on the FDA's Med Watch website. (POE 02004807-8)( Exh P-148). (POE 0207036-48; 02007049-50, P-151) This document included the October 19, 2005 FDA proposed revisions of ORTHO EVRA labeling with J&J's Counterproposal October 25, 2005 and rationale ((Exh P-146). (POE 02007049-50)(Exh P-151) J&J questioned the FDA's Counter-Proposal for ORTHO EVRA labeling received October 19, 2005.
64. In a J&J email dated November 3, 2005, there is a discussion of a 5:30 PM teleconference held between the FDA and J&J to discuss ORTHO EVRA's label supplement (POE 04891294). Dr. Florence Houn, Director, Office of Drug Evaluation III. led the FDA's discussion noting that no action would be taken by the FDA on the supplement of October 31, 2005. Dr. Houn explained that she had reviewed J&J's labeling proposal as negotiated with the Division and had found it unacceptable.
65. In a follow-up to the discussion, there is a J&J record dated November 17, 2005, documenting a November 1, 2005 teleconference which occurred between the FDA and J&J. (POE 04401871-3) (Exh 232). Dr. Florence Houn, FDA, began by acknowledging the efforts of both the FDA division and J&J to revise the labeling in good faith. She said upon further review by the FDA, the message in the bolded warning was unclear. She believed that the labeling needed to be renegotiated. Her directions that had come directly from FDA's Senior Management were that there was "a concern about the overall increased exposure from estrogen based on the data presented". The FDA also was concerned about the safety reports documenting thromboembolic events (TE) and she noted the Agency had already received 18 reports of death.
66. Dr. Houn said the FDA was confused as to how to interpret the different numbers with respect to AUC (for EE) from ORTHO EVRA now, "versus the data in the original application, versus the 35mcg OC". The relative risk [for thromboembolic (TE) events] appears to be about "2 times higher than with a 50- mcg OC versus those containing 30-40 mcg EE" and, in this context, the FDA was concerned about an increased risk possibly related to an increased EE exposure…" The FDA is particularly concerned with very young women on ORTHO EVRA with respect to PE, TE and possibly DVT? Or stroke? and death". Dr. Houn noted that she believed the indication should include a message for the prescriber to take into consideration the increased exposure to EE with ORTHO EVRA vs. OC.
67. On November 7, 2005, Dr. Houn explained to J&J that the FDA's goal of the label revision was now to balance the potential adverse consequences of increased exposure to estrogen with the consequences of increased risk of pregnancy. At the direction of FDA senior level management, the Agency had revised the section titled, Indications, to add a statement that would reflect this balance. In response to J&J's label change proposals, Dr. Monroe, FDA, indicated that the Agency believed the initial epidemiology study, referred to as the "SS. Jick study", should not be used in the label at this time. It was a single study and the data was still preliminary, with a second J&J epidemiology study to be provided by Ingenix (i3) later. The Agency was concerned that although the preliminary i3 data was available, it still had not yet been confirmed by a chart revi
ew. Dr. Houn explain
ed that the FDA could not make a final decision on label changes based on one report (SS Jick, et al.) and looked forward to reviewing additional data.
68. As was indicated by the FDA to J&J, the Agency wanted to focus the label changes solely on the ethinyl estradiol data. The FDA proposed addition of a statement that "systemic ethinyl estradiol exposure with ORTHO EVRA may approximate EE exposure attained with a 50 mcg pill" to increase the clarity of the significance and relate this to patient risk. The FDA removed any proposed reference to the data from the epidemiology study for the updated label. (See APPENDIX 6, page224-225)
69. J&J submitted NDA 21-018 S019 as a Changes Being Effected (CBE). The Labeling Supplement was dated November 8, 2005(POE.SP.000107–120)(Exh P-153). J&J proposed to accept some of the FDA's suggested changes of November 7, 2005 and provided the Division with its own new labeling proposal along with its rationale. According to J&J's points provided to the FDA for support of its new draft labeling:
ORTHO EVRA could not be compared to a 50 mcg OC;
there are differences between J&J and the Agency in interpretation of the rationale for the design of the BCDSP and i3 epidemiology studies;
the data from the BCDSP study did not demonstrate an increased thrombogenic risk with ORTHO EVRA in comparison with OC containing 35 mcg of EE;
and there is other epidemiology data on the relationship between estrogen dose and thrombogenic risk of OC.
70. As a result of J&J's decision regarding its labeling and negotiations with the FDA. the final November 2005 ORTHO EVRA label change did not include a statement warning that the systemic EE exposure of ORTHO EVRA for a woman was best compared to a 50mcg OC as had been proposed by FDA.
d. J&J's FAILURE TO ADEQUATELY DISCLOSE TO THE FDA AND PHYSICIANS USE OF A 60% CORRECTION FACTOR FOR NRGEEP-PHI-014
71. As supported by J&J's own internal documents and employee testimonies, the 20 mcg EE drug delivery on the original label was derived by J&J's PK expert, Dr. Larry Abrams, through his inappropriate use of a 60% correction factor in 1999 during calculation of serum EE values in J&J's phase I NRGEEP-PHI-014. NRGEEP-PHI-014 had been conducted as an optional J&J study, a Phase I bioavailability study, submitted to the FDA by J&J in its IND in 1999 to be considered by the FDA as part of NDA 21-180. As has been shown in J&J documents and testimonies, the serum EE values derived from NRGEEP-PHI-014 were false and misleading, reduced serum EE values by 40%, and were not consistent with current bioavailability studies in the medical literature used for measurement of EE in 1999. On July 26, 2000, Dr. Abrams also sent an internal J&J email stating to members of J&J that he did not want to have PHI-014 cited as the bioavailablity authority for the patch, but "I would rather cite published papers than have this reviewer nose around the PRI BA study data. "(Soons, Exhibit 972) By the unadjusted values of PHI-014, EE values of ORTHO EVRA clearly exceeded the delivered dose of a 35mcg OC. Dr. Abrams in an email of May 12, 2003 indicated that ORTHO TRI-CYCLEN EE was estimated to be 65.8% and 50.8% respectively.(POE 08372944) For a 35mcg OC, the systemic EE delivery would be 17.5 mcg EE, far less than ORTHO EVRA's "unadjusted" PHI-014 EE delivery.
72. Dr. Daniel Davis, FDA's Medical Officer in his November 6, 2001 review of NDA 21-180, wrote regarding the Agency's reliance on NRGEEP-PHI-014:
…At the inception of clinical development, it was anticipated that the 20cm2 patch would deliver a daily dose of 250ug 17d-NGM/25 ug EE. The results of Study PHI-014 subsequently showed that the daily dose delivered to the systemic circulation by the 20 cm2 patch is 150ug 17d-NGM/20ug EE. The results of the three pivotal contraceptive efficacy studies show that the EVRA patch provides contraceptive efficacy equivalent to that achieved with ORTHO-CYCLEN, a monophasic hormonal oral contraceptive containing 250 ug NGM and 35ug EE in each pill.
73. On June 21, 2007 J&J submitted, a June 15, 2007 report titled "Report Post-marketing Study of ORTHO EVRA and Levonorgestrel Oral Contraceptives Containing Hormonal Contraceptive with 30ug of EE in Relation to Non-Fatal Venous Thromboembolism, Ischemic Strokes and Myocardial Infarction," from the Boston Collaborative Drug Surveillance Program ("BCDSP"), Boston University School of Medicine (POE.SP.096016-81). It was an unpublished follow-up of the SS Jick epidemiology study (BCDSP). On the submitted FDA form, J&J indicated that the strength of the transdermal ORTHO EVRA patch was "20 mcg ethinyl estradiol", the incorrect value calculated with the use of the 60% correction factor for NRGEEP-PHI-014 (POE SP.096014).
74. As supported by the FDA's internal reviews of NDA 21-180, the NRGEEP-PHI-014 values from J&J's "labeling study" or "absolute bioavailability study", terms used by J&J internally, contributed to the FDA's eventual approval of the ORTHO EVRA NDA, of EMEA's approval of EVRA's MAA for marketing in Europe, of Health Canada's approval of EVRA's NDS for marketing in Canada. PHI-014 was the primary source of ORTHO EVRA's product description of systemic EE delivery in J&J's original drug label.
75. When J&J's management became widely advised of the inherent flaws in J&J's serum EE calculations in NRGEEP-PHI-014, as evidenced by a May 3, 2005 memorandum from Donald Heald, Ph.D. to Angie Giusti, Ph.D., Compound Development Team, three years after the April 30, 2002 commercial launch (POE 04818681-90), there was no recommendation made to "correct the error" by either informing the FDA or physicians of the correct EE delivery results. Instead, Dr. Heald wrote a "Problem Statement", with no indication of a J&J intent or duty to inform either the FDA or physicians of the error in EE drug delivery for ORTHO EVRA. J&J's management has appeared to continue to persist in not adequately informing the FDA, other regulatory health authorities, or physicians about the 1999 error in calculations. Such action by J&J demonstrates a continued disregard for women's safety, and ensuring its own compliance with the Act.
76. In an email dated January 15, 2004, Paul Soons, VP of Clinical Pharmacology, proposed that J&J repeat the PHI-014 bioavailability study when he wrote:
Just got a strange (great??) idea on solving the issue of which product to take as a reference production in a possible future (BE) study if we were to reformulate EVRA….why not take an IV comparison in such trial and really confirm what is in the label….releases 20 micrograms of EE to the blood stream per 24 hours….and…Css of EE during one week of patch wear is approximately 40-50pg/ml… (Exhibit 595).
77. In a June 13, 2006, email from Paul Soons, titled RE: EVRA-request for clarification from Health Canada, which he sent to various members of J&J, including Dr. Devineni and Dr. Heald, (POE 07597415-6), he discussed a Health Canada reviewer. Upon examining EVRA's NDS ,the reviewer questioned the NRGEEP-PHI-014 study report calculations. The reviewer specifically wanted to see J&J's calculations for NRGEEP-PHI-014 and how J&J derived the daily delivery rate numbers. J&J was also to provide its rationale for its choice to use clearance data and averaging of values derived from the buttock and abdomen. J&J was to submit the data back to Health Canada by June 23, 2006.
78. In a subsequent email dated September 13, 2006 from D. Devineni to various members of J&J's PRDUS, titled "MKM for the CDT", accompanied by a mass balance study-final-081206, J&J was again responding to Health Canada's concerns about EVRA and bioavailability information submitted by J&J. Health Canada now wanted J&J to complete a new PK study with EVRA within 4-6 weeks (POE 07582637-8). J&J's proposed protocol was titled: Mass Balance Study of Transdermal Contraceptive System (EVRA) of Norelgestromin and Ethinyl Estradiol in Healthy Female Volunteers (RWJ-10553/RWJ-1403-000) (POE 07582639-48).
79. Within several October 18-19, 2006 emails involving D. Devineni, J. Kiesewetter, as well as other members of J&J's PRDUS, titled NRGEEP-CON-1012 Post PRC Draft and draft responses to PRC minutes. The protocol was sent to Paul Soons for his approval. The proposed protocol is now a Phase 1 study, NRGEEP-CON-1012 titled Residual Analysis Study to Assess Release of Norelgestromin and Ethinyl Estradiol in Healthy Women Using a Transdermal Contraceptive System (POE 07573541-605), dated October 18, 2006. The objective of the study was to estimate the average daily amount of NGMN and EE released from a transdermal contraceptive patch system (EVRA) of NGMN and EE during application to the skin over 7-days. The secondary objective is to estimate the exposure of NGMN and EE following application of the EVRA transdermal patch to the abdomen and buttock. The EVRA patch was sold in Canada with a label stating it contained 6.0mg norelgestromin (MGMN) and 0.6mg of EE, unlike ORTHO EVRA sold in the United States.
d. FAILURE TO ADEQUATELY DISCLOSE NRGEEP-PHI-017 PK RESULTS
80. J&J continued to pursue ORTHO EVRA's NDA approval as if comparable to the EE delivery of a 35 mcg OC despite its knowledge of the complete results of its own pharmacokinetic (PK) phase 1 study, NRGEEP-PHI-017 in 1999. The results of NRGEEP-PHI-017 were available to J&J's management prior to NDA approval in November 2001, and clearly demonstrated measured increased systemic serum EE levels in women using ORTHO EVRA when directly compared to systemic EE delivery for a course of three 35 mcg COCs ( 2X increase EE to Triphasil, 3X increase EE to Mercilon, Alesse). Yet, the FDA was provided with an "Abbreviated Study Report" (POE 00017513-8) for PHI-017 for its NDA consideration with a later full report. Through 2007, the FDA has still failed to make a public reference to the data contained in PHI-017 in a manner that supports that the Agency has been fully advised of the 1999 complete results of ORTHO EVRA vs. OC data developed through NRGEEP-PHI-017.
81. The initial communications given to the FDA about PHI-017 prior to NDA approval was that the "pharmacokinetic data from this study are not being presented at this time." J&J provided the following discussion to the FDA regarding J&J's interpretations of PHI-017 data:
The pharmacokinetic data for EVRA in this study were consistent with or similar to pharmacokinetic data for EVRA in other Phase I studies in this program.
The safety evaluations for subjects in this study are presented in this Abbreviated Summary, with relevant listings appended (POE 00017514)).
82. An email was sent from Dr. Larry Abrams to Kei-Lai:
"The following is a rationale re: why study PHI-017 is not being included in the NDA. This paragraph appears in the abbreviated safety report. My preference for Section 6 is not to make a big deal about it. The protocol was submitted in the IND. Perhaps it could be a footnote to Attachment A and Table I that it is an ongoing study" (Exhibit 689)(POE 00347526).
83. An email was sent from Valerie Donnelly to L. Abrams regarding ORTHO EVRA's NDA Item 6, asking for his comments on the revised version of the letter to be sent from Dr. Polo to the FDA. The letter contained a footnote that:
PK parameter values for study NRGEEP-PHI-017 will not be provided in the original NDA. The data will be amended to the NDA as soon as available. Safety data for this study will he provided in the original NDA" (POE 00356653-56) (Exhibit 690).
84. Dr. Abrams wrote an email to Ceile Hedberg dated September 12, 2000 that:
…to include PHI-017 PK data in the NDA would mean also including it as a finished study in the Section 6 Human PK and BA ("bioavailability") NDA summary, the Clin Pharm NDA summary, and possible other critical NDA summary documents, etc." (POE 00356198-9)(Exhibit 692).
85. The complete NRGEEP-PHI-017 data combined with uncorrected NRGEEP-PHI-014 EE data, would have constituted significant scientific data essential for the agency's final risk versus benefit decision for approval or denial of approval of ORTHO EVRA intended for use in a healthy female population for prevention of unintended pregnancy. However, this data showing increased systemic EE delivery for PRTHP EVRA in relation to 35 mcg OCs does not appear to have been accurately and fully disclosed by J&J to the FDA prior to NDA approval in November 2001.
86. An example of the impact that knowledge of the complete NRGEEP-PHI-017 results could produce, in terms of evaluation of risk versus benefit, is evidenced by the inactions of Dr. Guillebaud. When EVRA's European Clinical Expert reviewer, Dr. John Guillebaud, a scientist who had already given EMEA his personal recommendation for EVRA's approval, received full disclosure of NRGEEP-PHI-017 results from Drs Creasy and Abrams, he was motivated to write a June 24, 2001 letter to Dr. Creasy urging: "Transparency with regards to the findings of PHI-017" to regulatory authorities. He also recommended that J&J publish the PHI-107 data (POE 01039910-11). Dr. Guillebaud wrote in his 2001 letter:
I write you formally in my capacity as invited Clinical Exert for the new combined hormonal contraceptive patch now known as EVRA, and I request that this letter be placed on your records as it will be in mine.
As you know, PHI-017 from a straightforward, uninterpreted, pharmacokinetic stand-point showed for EVRA a total AUC for ethinyl estradiol (EE) which was 2X Week 3 of Triphasil and 3X that of Alesse and Mercilon.
To conclude, I advise:
• Complete transparency with the regulator in regard to the findings of PHI-0 17, even if not specifically Questioned about the findings. This may lead to some agreed change to the labeling previously agreed by me, at the time of my Clinical Expert Report, when the findings of PHI-017 were not yet available.
• Publication of these findings of PHI-017, once fully analyzed statistically and put in context, so that international scientific community has the opportunity to evaluate its relevance or otherwise to the safety/efficacy/quality of life balance of this product.
87. Dr. Creasy testified that when Dr. Guillebaud reviewed the EVRA marketing application for the EMEA, he did not have the final results of PHI-017 to consider for his recommendation to the EMEA (Creasy, 07 p 271, 19-13). According to Dr. Creasy in 2007, he was unaware of J&J having taken any steps towards seeing that the results of PHI-017 were ever published (Creasy, 07 p 27, 1-11).
e. FULL DISCLOSURE OF NRGEEP-PHI-017 RESULTS ASSOCIATED WITH NEW ZEALAND'S 2004 DENIAL OF EVRA MARKETING APPROVAL
88. Another example of the impact of full disclosure of NRGEEP-PHI-017 in terms of the risk of the EVRA patch when compared to a 35 mcg OC occurred not with the FDA but with the New Zealand Medicines Assessment Advisory Committee (MAAC) in 2004. For J&J to obtain marketing approval to sell EVRA in New Zealand, the MAAC required J&J to fully disclose the results of NRGEEP-PHI-017. As a result of that disclosure, on December 2004, EVRA marketing approval was denied by the MAAC due to unacceptable risk based on NRGEEP-PHI-017 (POE 02271757).
f. DELAY IN ADEQUATE AND FULL DISCLOSURE TO THE FDA OF INCREASED SERUM EE DELIVERY OF ORTHO EVRA IN PRI/EDN-NED-I WHEN COMPARED TO A 35MCG OC
89. J&J continued to pursue commercial marketing of ORTHO EVRA to healthy US women despite management's awareness that there were increased serum EE values seen in post-approval study PRI/EDN-NED-I when compared to an 35 mcg OC. These findings were consistent with the results which had already been seen in the earlier 1999 Phase I NRGEEP-PHI-017. PRI/EDN-NED-I however was a post-approval study commitment made by J&J to the EMEA for continued EVRA marketing in Europe. At Dr. Creasy's August 3 , 2007, deposition, he was asked why J&J would have performed PRI/EDN-NED-I in 2002-2003 (comparing ORTHO EVRA to OC CILEST) for the EMEA when J&J already had in its possession the data from NRGEEP-PHI- 017. In terms of J&J's delay in full disclosure of ORTHO EVRA EE delivery when compared to a 35mcg, he responded that the FDA had "never requested" that an ORTHO EVRA comparison to an OC be performed. Unlike the FDA, the EMEA had requested that J&J conduct a post-approval OC comparison study (Creasy, 8/31/07 p 254, 17-15). Since the FDA had not requested an OC comparison study, the Agency would not have been looking for the completed data from PHI-017 for determination of NDA approval.
90. What Dr. Polo did not say, however was that as part of the NDA approval process, the FDA would anticipate that J&J would have fully and completely disclosed all data necessary for the Agency to consider in terms of risk versus benefit prior to approval of ORTHO EVRA. According to 21 CFR 312.32(b) IND Safety Reports-
"The sponsor shall promptly review all information relevant to the safety of the drug obtained or otherwise received by the sponsor from any source… including information derived from any clinical or epidemiological investigations.."
91. Per 21 CFR Sec 314.50, the summary of the NDA is to contain a summary of the application in enough detail that the reader may gain a good general understanding of the data and information in the application. There is to be a summary of the human pharmacokinetics and bioavailability section of the application as well as a summary of the clinical data section. As with the IND Safety Report, the NDA is to have a "description and analysis of any other data or information relevant to the evaluation of the safety and effectiveness of the drug product obtained or otherwise received by the applicant from any source". There is nothing that states that the FDA is not to be advised of safety and effectiveness information that it specifically requested from J&J, particularly when relevant to its safety and effectiveness risk/benefit determination of the Agency, such as PHI-017.
92. The NED-I study was conducted in The Netherlands, using U.S. ORTHO EVRA patches manufactured at Redwood City, CA. The NED-I patches were obtained from one of J&J's three manufacturing Process Validation lots, Lot 60L005. ORTHO EVRA drug delivery was compared to drug delivery in women taking a course of 35 mcg OC CLIEST. The NED-I study was completed in March 2003 with the report finalized on October 28, 2003.
93. A series of emails (POE 02046539-42)( Exhibit 183) were sent between Patrick Caubel, Larry Abrams, and Daniele Scarcella regarding the NED-I interim draft report. Dr. Caubel said that "from a benefit-risk perspective, he wanted to avoid the take home message from this preliminary report that EVRA is equivalent to a 43 mcg OC". He also was highlighting that the mean AUC values were abnormally high compared to those reported in the 021. regardless of the manufacturer. Daniela Scarcella did not want to introduce more "issues" into this I week data discussion and preferred to leave this discussion for later. Both she and Dr. Abrams proposed a "wait for the rest of the data" approach.
94. Dr. Caubel however continued the discussion with the following clinical concerns:
Let me stress again that the implications of these findings go beyond this study report. With an EE AUC of 9723 pg/h/ml we are clearly out of the SD interval of the AUC reported in the USPI (6796*2673 for cycle 1-week 1). The estrogenic exposure is a critical piece of the clinical benefit/risk and safety profile of an EP contraceptive combination, and I would like to make sure that why these findings from study PRI/EDN-NED-I do not reflect the PK profile of the currently marketed product is discussed as some point.
May I remind you that in the introduction of the pharmaco-epidemiology VTE protocol currently under review by the EMEA we are claiming that steady state plasma levels are comparable between EVRA and CILEST, and the product was based on this assumption.
95. After the FDA's approval of the ORTHO EVRA NDA in November 2001, the Agency conducted its first internal Adverse Experience Reports (AERs) safety review of ORTHO EVRA in 2003. The study was done by the Division of Drug Risk and Evalution (DDRE) at the request of the Division of Reproductive and Urological Products (DRUP). The FDA had gained awareness of a trend in increased reporting of serious adverse events (SAEs) and pregnancy with ORTHO EVRA as of 2003, one year after commercial launch.
96. On January 19, 2004, J&J submitted its Annual Report for NDA 21-180 for the period November 21, 2002 through November 20, 2003. It was a one volume report (POE 00088791-63). The coversheet was brief without any mention of significant safety issues occurring during the time period. It was submitted by Patricia Capaccione, R. Ph, Senior Associate, Regulatory Affairs, Global marketed products. Under the Summary of Significant New Information for the reporting period, J&J indicated to the FDA that there was no significant new information to report (POE 00088796). On page 127, there is a listing of the published clinical reports that had become available during the period. The FDA was told that copies of the full reports would be provided to the Division upon its request.
97. The report contained the following paragraph with nothing to call attention to the Agency reviewer of the type of safety and performance information contained in the described post-approval "NED-I study". The PRI/EDN-NED-I study had been initiated July 2, 2002 and completed March 10, 2003 as a post-approval study for EVRA to remain marketed in Europe. J&J wrote in its Annual Report to the NDA the following about PRI/EDN-NED-I:
The following clinical study was required as a post approval commitment for the registration of this product in the European Union. The commitment required that we conduct a study in which the pharmacokinetics of CILEST and EVRA are compared in a cross-over design study. A copy of the final clinical study report is appended to this section. There were no other unpublished reports that became available during the reporting period.
PRI/EDN-NED- I RWJ- 10533-097/RWJ- 1403-000(EDMS-USRA-84464:2.0)
A Comparative Pharmacokinetic Study of EVRA and CILEST in Healthy Female Volunteers.
98. On May 14, 2004, J&J provided the FDA with a Briefing Package in Support of a Meeting Scheduled June 29th, 2004 (POE 00095727-842; POE 00096145-50), POE 00096405, POE 00096408). The purpose of the meeting was to discuss three questions with the FDA regarding Pharmacokinetics/Risk Management/Chemistry& Manufacturing & Controls and to obtain FDA's agreement on J&J's proposed evaluation plan. There was no discussion of the recent publicity regarding a young woman's death in New York in April or a story in the New York Post nor discussion of the safety concerns of the NED-I results . Instead, J&J provided the FDA with the following three questions as the agenda or purpose of the meeting to be discussed with the FDA (*underlining added for emphasis) (POE 00095727-842: POE 00096145-50; POE 00096405; POE 00096408)(APPENDIX 6 ¶ 321- 322):
I. Recent clinical studies evaluated pharmacokinetics of 4 ORTHO EVRA lots (2 commercial lots and 2 process development lots) during the post-approval period. Alth
ough the
re was substantial overlap in individual subject data, the mean I-week AUC values for ethinyl estradiol (EE) in these post-approval studies ranged from 9,409 to 10,061 pg.h/mL, whereas mean AUC values during the clinical development program 3 to 8 years ago ranged from 5,750 to 9,101 pg.h/mL. Unfortunately, no direct head-to-head trials can be conducted, as retained samples from the clinical lots have expired. We are conducting a more comprehensive PK study (protocol outline in Appendix 1), which will examine a more representative sample of commercial lots. We believe that this approach will provide more comprehensive information to determine the average EE exposure from the current commercial lots of ORTHO EVRA. Does the Agency agree?
2. Post-marketing surveillance has collected a large volume of adverse events in patients using ORTHO EVRA. The company has implemented specialized questionnaires to obtain more details about serious thromboembolic events and pregnancy reports. We are also conducting a comparison of adverse events reported with ORTHO EVRA to other contraceptives. Our analysis indicates that no new events, trends, or significant risks have emerged from the ongoing review of safety with ORTHO EVRA and we believe that our current product labeling is appropriate. Does the Agency agree with this assessment?
3. All released lots have been within specifications. However, the Company has also undertaken a CMC investigation that is ongoing, which includes: review of changes to the manufacturing procedure from Phase 3 to present; review of batch records of lots associated with spontaneous post marketing reports of venous thromboembolisms, and implementation of a design of experiment (DOE) to further characterize the factors influencing product performance. Details of these studies are presented in the briefing package. Does the Agency agree with these evaluations?
99. The FDA was told that since the FDA's approval of ORTHO EVRA, J&J had conducted three studies between 2001 and 2003 that included pharmacokinetic (PK) analyses: PRI-EDN-NED-I; NRGEEP-POI-121; and NRGEEP-PHI-023. These studies evaluated four ORTHO EVRA lots (2 commercial * 2 process development) made at the Redwood City, CA manufacturing site, but did not include any comparison to historical patches from the clinical development program. To date, more than 150 lots have been manufactured and released by J&J. reportedly with all meeting specifications. There was a footnote that PRI-EDN-NED-I had already been sent to the Agency on January 19. 2004 (i.e. NDA Annual Report); NRGEEP-PHI-023 had been sent to the FDA on October 3. 2002 and NRGEEP-POI-121 would be sent in J&J's upcoming IND Annual Report. Dr. Heald, who had formerly worked at the FDA, indicated at his deposition that "Annual Reports" like the January 2004 Annual Report with the PRI/END-NED-I study results are considered routine submissions at the FDA. Because of that, and the June 2004 meeting, "that's why we decided that we better contact them and -with a meeting request just so it (*NED-I) didn't get lost in the shuffle" (Heald 2/1/07, p. 92, 1.1-11).
100. The FDA first became adequately aware of the results of PRI/EDN-NED-I, as a result of J&J's May, 2004 Briefing Package. PRI/EDN-NED-I results showed the FDA a 60% increase in systemic EE delivery to women when compared to EE delivery of 35 mcg OC CILEST. As a result of the Agency's learning of the NED-I results, as well as the FDA's concerns about increased reporting of serious adverse events, death and unplanned pregnancy with ORTHO EVRA, the FDA requested J&J to revise the label to provide physicians and patients with adequate warnings about the risks of ORTHO EVRA and EE delivery.
101. Also in May 2004, J&J, as a post-approval requirement for the EMEA, began conducting NRGEEP-POI-1025 a EMEA post-approval study conducted in Antwerpe, Belgium using LTS manufactured EVRA compared to a 35 mcg CILEST OC. The study was completed on October 8. 2004 and the report finalized on April 21, 2005. The study also compared German manufactured LTS EVRA successfully to serum EE release data of the historical Phase 3 clinical lot (Lot 10607). As a result of the successful outcomes for EVRA, the EMEA would make a distinction between the risk of EE reported for US ORTHO EVRA versus European EVRA.
g. FAILURE TO ACCURATELY DISCLOSE THE LACK OF THERAPEUTIC EQUIVALENCE OF FOUR ANATOMIC SITES
102. J&J's NDA, provided phase 1 studies, NRGEEP-PHI-004 and NRGEEP-PHI-013, to the FDA to support therapeutic equivalence of the four potential body sites for application of the ORTHO EVRA patch for 7-day delivery of combination therapy. The only J&J study where placement on the abdomen's systemic EE drug delivery measured was higher than the buttock was NRGEEP-PHI-013. In all other ORTHO EVRA studies, the abdomen always produced a 25% lower systemic level of EE. Because of the data supplied by J&J in NRGEEP-PHI-013, the FDA was allowed to incorrectly conclude that all four anatomic sites were "equivalent".
103. According to Dr. Abrams, at his deposition of June 2007, J&J would learn that application of the patch to the abdomen would produce a 25% reduction in measured serum EE concentrations. Dr. Abrams, J&J's PK expert through the development and launch of ORTHO EVRA until his retirement in 2004, did not consider the abdomen as therapeutically bioequivalent to the other 3 body sites (Abrams, 6/7/2007, p.108, 1.20-22)). This was in contrast to the information he provided to the FDA in the IND and NDA for PHI-004 and PHI-013.
OPINION #2:
104. J&J disregarded its duty and obligations as a pharmaceutical manufacturer, ignored forty-years of oral contraceptive history and science, failed to utilize good scientific methods and appropriate evaluation of its pre-clinical safety signals for design, development, testing and marketing of ORTHO EVRA. J&J's scientists in 1996, had calculated that a 20cm2 patch should be loaded with 60 mcg of EE to deliver a nominal 20 mcg EE per day based on EE skin flux of 0.04 mcg per cm2 per hour. The EE delivery values had been based on PK characteristics of ORTHO CYCLEN OC and were aimed at obtaining a steady state range of 40-50 pg/ml (Audett, 7/23/07, p 254, I. 18-20). However, J&J's development team subsequently made a determination to increase patch EE adhesive loading to 75 mcg to increase the delivery of EE. After increasing the systemic EE delivery for the patch, J&J disregarded the safety signal of increased reporting of thrombotic events during the Phase 3 clinical trials as well as disregarded a February 20, 1998 Leiden-Site report to Dr. Shangold of unacceptable estrogenic side effects suggesting high systemic EE delivery from J&J's clinical investigator Dr. Peter Van der Ark (POE 04854048). J&J also disregarded the coagulation results of CONT-006 (APPENDIX 6 30,109,178,198, 222-225). Furthermore, during its pre-NDA negotiations with the FDA, J&J failed to adequately inform the FDA that as early as 1998 it had been aware of safety signals of increased reports of pulmonary emboli greater than anticipated for a 35 mcg OC. After obtaining the FDA's NDA approval, J&J continued to elect not to provide its own clinical experts , key opinion leaders ("KOL"), physicians and patients with adequate warnings about increased systemic EE delivery and reporting of VTE in its clinical trials in its original ORTHO EVRA label, and communications (APPENDIX 6 ¶ 283-288). J&J was aware that, unlike ORTHO EVRA being sold in the United States, J&J's EVRA patch sold outside the US was loaded with 60mcg of EE. EVRA, and also unlike ORTHO EVRA, had been able to support EE bioavailability to a 35 mcg CILEST OC (NRGEEP-POI-1025) as well as bioavailability to historical data for J&J's clinical trial lot 10607. J&J's actions regarding its follow-up of the safety of increased EE delivery of ORTHO EVRA has shown a disregard for ensuring patient safety and a disregard for its own duty to sell safe and effective products. Such actions by J&J have directly contributed to the injuries and deaths of healthy young women who elected to use ORTHO EVRA for prevention of unintended pregnancy.
ADDITIONAL SUPPORT FOR OPINION #2
105. J&J has repeatedly demonstrated its own knowledge and awareness of the increased risks of VTE associated with increasing serum levels of EE in COCs and ORTHO EVRA. J&J as a pharmaceutical manufacturer should be aware of its responsibilities and duties under the Act to fully and truthfully communicate risks of new products to the FDA, physicians and patients as well as the risk for unanticipated pregnancy (reduced efficacy).
106. In 1988, J&J, as well as the rest of the United States oral contraceptive industry voluntarily withdrew from the market all COCs with EE greater than 50 mcg. A COC with 50mcg of EE delivers approximately 25mcg of EE to the systemic circulation after liver first-pass. This industry reduction in delivered systemic EE was done based on FDA, physician, industry and public safety concern about unacceptable risks of VTE and death with no additional gain in efficacy ( prevention of unintended pregnancy).
a. J&J WAS AWARE IN OFFICIAL COMMENTS TO THE FDA REGARDING THE AGENCY'S 1999 DRAFT LABELING GUIDANCE OF RISKS OF INCREASED SERUM EE DELIVERY AND ITS RESPONSIBILITY TO PROVIDE ADEQUATE WARNINGS
107. In a J&J September 29, 2000 letter (POE 00506689-720), J&J commented to the FDA regarding the Agency's 1999 proposed draft for Combined Oral Contraceptive- Labeling for Healthcare Providers and Patients Docket 00D-1350 (See APPENDIX 6). J&J wrote regarding its own responsibilities: "As a commercial marketer, we believe it is our responsibility to objectively relate the risks of oral contraceptive use so as to fairly and adequately inform those who prescribe our product"…. "New and novel estrogens and progestins could have unique characteristics which would affect a risk/benefit assessment" (POE 00506694). J&J also wrote the following proposed label Warnings:
COC use is associated with an increase in the incidence of cardiovascular disease. The degree of risk appears to be related to the steroid hormone dosage, especially in products containing estrogen dosages greater than 35mcg(1). Minimizing exposure to estrogen and progestogen is in keeping with good principles of therapeutics. For any particular estrogen/progestogen combination, the dosage regimen prescribed should be one that contains the least amount of estrogen and progestogen that is compatible with a low failure rate and the needs of the individual patient.
108. On page 11, J&J's point #26, (POE 00506699):
Current clinical recommendation involves the use of lower estrogen dose formulations (containing 35 mcg or less of ethinyl estradiol) and careful consideration of risk factors.
109. One of J&J's own clinical officers, Dr. K. LaGuardia, has testified in 2007 that she was aware of the increased VTE risk associated with prescription of OCs with EE greater than 35 mcg (K. LaGuardia, 01/17/07 p 240-244).
b. IN A JULY 1, 2000 EMAIL, J&J ASKS A QUESTION AS TO HOW MUCH IT WOULD COST TO REDO ALL ORTHO EVRA CLINCIAL STUDIES.
110. A July 1, 2000 email was sent from Jill Ferentz, PRI to Y. Williams, L. Abrams, et al. titled : Costs to Redo Clinical Trials asking the development team how much it would cost to start all over again with EVRA and to repeat the clinical studies The information was to be presented at a GOC meeting July 13, 2000 (POE 00354674-5). The estimate to repeat all EVRA clinical trials again was $2.8 million. The clinical trials were not repeated and the NDA application was submitted to the FDA (APPENDIX 61168).
c. J&J's DELAY IN REPORTING TO THE FDA INCREASED RISKS OF PULMONARY EMBOLI OCCURRING IN J&J's CLINCIAL TRIALS
111. Despite J&J's past actions, its knowledge of risks and knowledge of requirements of the Act, J&J continued to pursue ORTHO EVRA marketing approval from the FDA while continuing to claim that its risks were comparable to a 35mcg OC. J&J made these claims comparing ORTHO EVRA to a 35 mcg OC despite its own awareness that it already had received increased reports of VTE during ORTHO EVRA's Phase 3 clinical trials in 1998 (CONT-002, CONT-003) (POE 05445264; POE 05443750). As supported by the adverse experience reports (AER) documents, J&J was aware in 1998 that ORTHO EVRA had been associated with an increased VTE risk of "118/100,000 woman years" (FDA's Medical Officer, Dr. Davis, NDA review; Dr. John Collins, Health Canada NDS review(POE 00460401); J&J's Consultant Heidi Jolson, M.D. September 2001 (POE 00413321-31).) (APPENDIX 6¶ 123,257,151-164).
112. However, J&J did not inform the FDA of the increased VTE risk during its July 1999 pre-NDA negotiations in its briefing documents, tabular summaries of reported adverse events from the IND 50,488 or during oral presentations made to FDA by Drs Shangold and Creasy(APPENDIX 6 ¶144-149).
1. NRGEEP-CONT-002; PRI/EDN-INT-2- SUBJECT# 21022- AER JULY 7, 1998- PULMONARY
EMBOLISM IND 50,488
113. J&J completed an AER (POE 05443750) initial report for IND 50,488, on the same day as its submission of its IND Protocol amendment for a serious adverse event with Subject 1022, a 34 year old woman enrolled in an international ORTHO EVRA study (NRGEEP-CONT-002, PRI/EDN-INT-2). The woman had elective plastic surgery on June 5, 1998. She was subsequently hospitalized on June 24, 1998 for pulmonary embolism and arterial thrombosis in the lower legs. This was the initial report on MedWatch Form 3500 A of a June 24. 1998 event and filed by Lynn Hopkins, MD, Clinical Studies LTD, Winter Park, FL. The investigator's opinion was that the event was possibly related to the study therapy and the surgery was definitely a contributing factor. Therapy dates were March 20, 1998 through June 4, 1998. The report had been received by J&J on July 7, 1998. The complaint was assigned manufacturer report number: 980707-107012420.
114. There is a second report for this same patient, a follow-up report dated February 2, 2000. The original event date was June 24, 1998 (POE 00014482-3). The date the manufacturer was informed was listed as February 2000 for a follow-up report. The event report is the same, as well as the manufacturer number (APPENDIX 6 ¶ 124-128). There is follow-up information supplied:
The initial pre-study PE showed no cardiovascular abnormalities, including any abnormalities of the peripheral circulation or extremities. All entry laboratory were within normal limits. An entry was made by the study coordinator stating pulmonary emboli and blood clots to the lower extremities. This subject had never been evaluated by the PI (or sub-I) regarding the above described event. The hospital records have not been obtained and the subject was lost to follow-up. Based on the available data it is most unlikely there were emboli to the lower extremities, i.e. of arterial origin.
2. NRGEEP-CONT-003; SUBJECT# 1181- AER SEPTEMBER 29, 1998, PULMONARY EMBOLISM
IND 50,488
115. J&J completed a second Form 3500A MedWatch report initial report for its IND 50.488. This time J&J described a serious thrombotic adverse event for Subject #1181, a 28 year old woman South African enrolled in an International ORTHO EVRA study (NRGEEP-CONT-003, PRI/EDN-INT-3) (APPENDIX 6 ¶ 124-128) (POE 05445264). This was the initial report made by J&J on Form 3500 A of an August 26, 1998 adverse event (*date of diagnosis of pleurisy), pleurisy and pulmonary embolism, filed by Shirley Mortimer, CRA, JANSSEN-CILAG. South Africa. The investigator's opinion was that event was probably related to the study therapy.
116. Subject #1181's therapy date with the patch began December 16, 1997. She was diagnosed with pleuritis on August 26, 1998 by x-ray. On September 18, 1998, she was hospitalized for a blood clot in the lung and discharged September 24, 1998 on Warfarin (Coumadin) and Clexane (enoxaparin) 40 injections. The study therapy (patch) was permanently discontinued on September 18, 1998. The report had been listed as received by J&J on September 29, 1998 the same date of the filing of its initial report. The complaint was assigned manufacturer report number: 980929-088013722.
117. A follow-up report for manufacturer report number: 980929-088013722 was filed by J&J on May 25, 1999 for an adverse event it now listed as occurring September 18, 1998( *not the date of initial diagnosis of pleurisy, but the later date of the diagnosis of her pulmonary embolism). J&J supplied the additional information received from its affiliate on November 30, 1998. The event was judged as resolved without sequelae on September 22, 1998. In the investigator's opinion, the event was changed from probably to "possibly" related to the study therapy. (POE 00014211-13).
118. The FDA was first advised by J&J of its "safety concern" of increased reporting of thrombotic events not in 1998 during ORTHO EVRA clinical trials, but on September 3, 1999 through an IND Safety Report following a prior September 3, 1999 telephone conversation held between Dr. Shangold and Dr. Rarrick, FDA (Polo, August 9, 2007, p 222,1.12-16; p224, 1. 11-14). J&J stated in its letter to the FDA that it recognized there was a "possible increase over the expected number of thromboembolic events" observed in the "relatively small group of women from three studies conducted both domestically and ex- US." There were six cases of adverse events classified as thrombotic and/or embolic in nature in patients using the study patch, and one in a subject using a comparator oral compound. All studies had already been completed and no additional patients were still using the patches. (POE 00014140-1). There had been a report of thrombophlebitis in the Phase 1 clinical trial in a woman using the smaller 15 cm2 patch. "Thromboembolic events" occurring during the ORTHO EVRA clinical trial, including pulmonary embolism would be relevant to the FDA's determination of "safety" of the ORTHO EVRA patch, which is a new unique form of hormonal contraceptive product (APPENDIX 6 ¶ 151-164).
119. J&J is required to make timely reports to the FDA for pending INDs. J&J is to provide the FDA with all information relevant to safety of the ORTHO EVRA patch as required under Sec. 312.32. Per 21 CFR §312.32, IND Safety Reports are to be written reports a sponsor uses to notify the FDA and all participating investigators of any adverse experience (i.e. safety) associated with the use of the drug (ORTHO EVRA) that is both serious (Pulmonary Embolism) and unexpected; or any finding from tests in laboratory animals that suggests a significant risk for human subjects. Each notification shall be made as soon as possible and in no event later than 15 calendar days after the sponsor's initial report of the information. Each safety report is to be submitted on the FDA's Form 3500A or in a narrative format. The sponsor is to notify the FDA by telephone or by facsimile of any unexpected fatal or life-threatening experience associated with the use of the drug as soon as possible but in no event later than 7 calendar days after the sponsor's initial receipt of the information. For follow-up, the sponsor shall promptly investigate all safety information received by it. The follow-up shall be submitted as soon as the relevant information is available.
120. According to Dr. Polo, notification of the FDA of thromboembolic events in the clinical trial had been the responsibility of ORTHO EVRA's Clinical group with Drs. Shangold and Creasy. When Dr. Polo was asked why it appears to have taken so long for J&J to have notified the FDA of thromboemblic events and PEs occurring in the completed clinical trials, he indicated simply that "you would have to ask Dr. Shangold and Dr. Creasy" (Polo, August 9, 2007. p201, 1, 4- 10).
121. J&J however, after the September 3, 1999 notification, waited an additional six months to submit further information in a follow-up letter to the FDA for its IND safety report on March 13, 2000. On March 13, 2000, Dr. Polo's signature once again appeared on J&J's IND Safety Report follow-up letter but now was sent to Dr. Susan Allen, Acting Director DRURP. The letter discussed the same 6 cases mentioned in J&J's September 3, 1999 letter to Dr. Rarrick. According to Dr. Polo, the cases with reports of superficial phlebitis without appropriate treatment for DVT/PE were assigned by J&J to a final disposition as "no DVT/PE" in the absence of an additional confirmatory test. Therefore, on March 13, 2000, the FDA was officially advised by J&J that it had filed two AERS reporting women with PEs in IND 50,488 for AERs reported to J&J in July and September of 1998 (APPENDIX 6 ¶ 151-149).
122. Marianne Mann, M.D., Deputy Director, Division of Reproductive and Urologic Drug Products (DRUP), CDER, FDA sent J&J an April 17, 2000 follow-up letter (POE 00014146) (Exhibit 804) to Dr. Polo and J&J regarding its March 13, 2000 submission. Her initial question to J&J was for it to provide the FDA an explanation for a 6-month time lag between the initial safety report (N-078) and the follow-up safety report. Her second request was that J&J provide the FDA the CRFs (case report forms) for the seven listed subjects provided in Attachment A of the submission. Thirdly, the FDA requested that it receive hospital discharge summaries, operative notes, and detailed lab results for both of the PE subjects, subjects 1181 and 21022. For subject 21022, the FDA also requested additional information about anesthesia and the last dose of EVRA. The FDA wanted further information concerning an additional case from the phase 1 study with the 15cm2 patch. The FDA wanted information about the patch application sites used for all the eight subjects. Finally, the FDA requested clarification of the different VTE rates and confidence intervals, with the VTE calculation to include subject 21022 (APPENDIX 6 ¶ 162).
123. J&J's Drs Creasey and Shangold, the clinical leads for the ORTHO EVRA Team, doggedly adhered to a derived venous thrombosis event rate (VTE) based on one PE (#1181) in the clinical trial or risk of 59/100,000 women years. As indicated by Dr. Polo, (Polo, 8/9/07, p202, 1. 11-16) apparently Drs Creasy and Shangold discounted the second PE because it was considered a protocol violation, therefore not applicable to ORTHO EVRA's VTE calculation (APPENDIX 6 ¶ 178).
124. In contrast to J&J's internal medical experts' opinion, the FDA's medical reviewer, Dr. Davis, did not discount the reporting of the second PE, #22101. Dr. Davis' regulatory counter-part in Canada, Canadian Medical reviewer, Dr. John Collins, in his review of the EVRA marketing application (aka NDS- New Drug Submission) for Janssen-Ortho also did not elect to discount the second PE (POE 00460401). J&J's former FDA consultant, Heidi Jolson, MD, advised J&J th
at the FDA would c
ount both of the PEs from the clinical trial (See Meeting with Heidi Jolson, M.D.-September 26, 2001 APPENDIX 6 ¶ 197-200). J&J's clinical investigator in its Phase 3 CONT-002 clinical study report caring for patient 21022 indicated that ORTHO EVRA was "possibly" related to the PE (POE 004861).
125. The FDA calculated two reports of PEs in the J&J clinical trials, producing a VTE incidence rate of 118/100,000, above the rate anticipated for a 35 mcg oral contraceptive (APPENDIX 6 ¶ 123,257). Despite the FDA's calculations, the original ORTHO EVRA label provided to physicians and patients had no VTE rate in terms of comparison to VTE risk for COCs. J&J elected to describe to physicians that two women in the clinical investigations had experienced a pulmonary emboli.
d. BY 2003, J&J HAD BEGUN CONSIDERING DEVELOPMENT OF A LOW DOSE ESTROGEN PATCH BASED ON "SAFETY"
126. There is a June 17, 2003 email from Dr. Abrams responding to an earlier email from Mr. Audett regarding his request for a rationale for development of a new low dose EE patch. Dr. Abrams wrote back that lower EE dosing is a "safety issue". He also indicated that over the years, both the progestin and EE of the combination OCs have been reduced but still have to be kept at efficacious levels. The EE has to be "high enough to maintain adequate cycle control" (reduce tendency for bleeding and spotting, etc. POE 08307254)(Exhibit 676) (APPENDIX 6 ¶ 262).
127. During the August 18, 2005 FDA and J&J meeting, as noted in J&J's minutes (POE 05059202-5), the FDA's Dr. Monroe, when looking at J&J's proposed label changes, stated that it seemed that ORTHO EVRA gave higher concentration values of EE than a 35 mcg pill and "perhaps as high as a 50mcg pill of higher". Since increased exposure to EE is related to increased thrombotic events, he thought that the labeling needed to focus on this information. Additionally, he felt there should be a statement that thinner women were at higher risk in using the product.(APPENDIX 6¶ 443,469,474)
e. J&J HAD INCREASED ORTHO EVRA's ESTROGEN DELIVERY EMPIRICALLY WITHOUT GOOD SCIENTIFIC SUPPORT, DISREGARDING THE POTENTIAL RISKS, AND FAILING TO ADEQUATELY MONITOR AND UPDATE LABELING
128. The ORTHO EVRA patch technology was licensed by J&J from Cygnus in 1994. The patent for ORTHO /EVRA extends worldwide through 2015 (POE 00204229). Cygnus is the holder of two patents for the patch technology, including US Patent 5,972,377 (assigned October 26, 1999, inventors Jannan Jona, Jay Audett, and Noel Singh). Ortho-McNeill Pharmaceutical, Inc. ("J&J") is the patent holder for the "Method of Manufacturing the ORTHO EVRA patch" (U.S. Patent No. 6,071,531 assigned June 6, 2000, inventors Jannan Jona, Jay Audett, and Noel Singh). J&J is the US patent holder for other transdermal drug delivery technology besides the ORTHO EVRA patch.
129. According to Cygnus' product development specifications (POE 03325516) the amount of ethinyl estradiol (EE) to be delivered from the patch is 0.75mg. The EE "range" considered as suitable for the patch system could be [0.60-0.90 mg/patch]. According to Mr. Jay Audett, that number and range came directly from J&J's analysis of EE delivery of 35 mcg ORTHO CYCLEN.
130. Dr. Gary Shangold, J&J Clinical officer and OB/GYN ( Shangold, 6/5/ 2007, p. 74-5, I. 17-3), provided his own clinical understanding for J&J's rationale for developing a 7-day contraceptive patch. The rationale included improved patient compliance, reduction in steroid dose with reduction of liver first pass, potential reduction in thromboembolic risks due to a lower estrogen dose and potential elimination of side effects, such as nausea, commonly associated with oral contraception steroids.
131. According to the deposition testimony of Mr. Jay Audett, an individual actively involved with the design of the prototype patches and listed on the U.S. Patents as one of the inventors, he designed the amount of ethinyl estrogen (EE) and norgestromin (17d-NGM or NGMN) to be delivered by a patch based on hormone delivery specifications he received directly from J&J's experts. The Cygnus Contraceptive Transdermal System (CTS) in the pre-Phase 1 clinical study showed that a " 20 cm 2 patch could deliver a nominal dose of 165 mcg 17d-NGM and 24 mcg EE per day" (Audett, 7/23/07, p 50-51 1. 20-1). The target skin flux used by Cygnus was 0.04 mcg per cm2 per hour (Audett, 7/23/07, p 254, 1. 18-20), and average skin flux value thought to be therapeutically equivalent to the amount of drugs delivered by a course of Ortho's oral contraceptives. The calculation of the Area Under the Curve (AUC) was not part of his expertise, but had been the domain of J & J's PK expert Dr. Larry Abrams (Audett, 7/23/07, p. 182-4). The flux target for EE was based on the established target dose from PK characteristics of an ORTHO TRI-CYCLEN obtained by J&J. (Audett, 7/23/07 p 185, I. 13-16)
132. Dr. Larry Abrams, retired from J&J since 2004 and now a private consultant, testified (Abrams 6/25/07 p 46, 1.13-21), when asked about the origin of the initial 60 mcg EE used for the prototype patch indicted that the amount had been derived from numerous calculations done primarily by scientists at "Cygnus" with some assistance from him. Dr. Abrams. J & J's PK expert throughout the ORTHO EVRA development project received his clinical support from Dr. Gary Shangold and Dr. Creasy, J & J's clinical experts for the ORHTO EVRA project.
133. Dr. Shangold (Shangold, 6/5/07, p 78 I. 2-8), who's knowledge of the ORTHO EVRA patch extended from development through 2002, indicted that, for a 35 mcg EE COC, despite documents stating 50% liver first pass of oral EE (i.e. 17.5mcg), he calculated perhaps 20 and 25 mcg of estrogen (55 to 70% of the 35ug) would actually survive liver first pass. He also personally thought that the threshold for increased venous thromboembolic risk ("VTE") for EE and OCs for women fell somewhere between 50mcg and 75 mcg /day EE with a value for increased risk never established (Shangold, 6/5/07, p 82 1. 1-25.). When asked about whether it was standard manner of care for physicians to prescribe the lowest dose of estrogen for a patient, Dr. Shangold (Shangold, 6/5/07, p 85, 1.17-25) responded that he thought some physicians would take that approach, but he was unaware of any policy statement from ACOG or the FDA that would dictate that every patient should be started on OCs at 15-20 mcg estrogen. (Shangold, 6/5/07, p 94, 1.1-11). He indicated that he saw no differences in risk of VTE for an estrogen dose between 15 or 20 mcg and even 50 mcg. He was aware of only one paper to support that there was a difference between VTE events and 50 mcg and above 50mcg. He continued (Shangold, 6/5/07, p 94, 1. 8- 15) "There's never been a demonstration of any continued increase in risk as one goes from 50 to 60 to 70 and 75 mcg.", "I think the reason for that is I don't believe there have ever been any products that were commercialized that contained doses of estrogen between 50 and 75 mcg. So there isn't any data."
134. Mr. Audett does not recall having been informed by members of J&J that there was a dose of EE delivery that would not be considered acceptable or unsafe for delivery from the patch.
d. J&J'S 2003 DECISION NOT TO FUND THROMBOSIS RISK FACTOR STUDY FOR ORTHO EVRA
135. A January 31, 2003 Ortho-McNeill inter-office memo contains meeting minutes sent from John LoCoco regarding Women's Health Care- Phase IV Investigator Proposal Meeting. A study proposal requesting funding from J&J was submitted by Jane Lowell, MD-
"Effects of an oral compared to a patch contraceptive on thrombosis risk factors" (FYI: I have included a copy of an abstract by C. Kluft on ORTHO EVRA vs OCs on coagulation parameters).
The decision was made not to fund this study for the following listed reasons: Risk that ORTHO EVRA may be the same or worse than Ortho-Cyclen; too high a chance that study may not produce a positive results for EVRA.
Instead of receiving funding for her research, Dr. Lowell was provided an abstract by C. Kluft on ORTHO EVRA vs. OCs on coagulation parameters. (POE 00328435-6) (APPENDIX 6 ¶ 260).
OPINION #3:
136. According to J&J's employee testimonies and internal documents, J&J failed to act as a responsible manufacturer when it aggressively marketed ORTHO EVRA as a contraceptive patch to healthy young women while internal management was aware it had failed to develop the necessary manufacturing release specifications and process controls to be able to reliably ensure safe and effective performance of the patch when used by a woman for prevention of unintended pregnancy (APPENDIX 6 MANUFACTURING 1 1-83).
# I. FAILURE of J&J TO DEVELOP ADEQUATE & CLINICALLY MEANINGFUL RELEASE SPECIFICATIONS for COMMERCIAL PRODUCT ("NO IN VITRO TO IN VIVO CORRELATION"):
Prior to NDA approval on November 2001, J&J failed to develop meaningful in vitro/in vivo correlation (IVIVC) specifications for determining the appropriate release criteria for a patch. J&J and Cygnus had initially used Franz cell skin flux testing to monitor potential drug delivery during patch development, design and clinical trials (Audett, 07/24/07 p.351, 1.23). Franz cell skin flux testing was subsequently deleted from commercial manufacturing process controls. J&J had been able to obtain NDA approval with minimal testing and data based on J&J's ability to reference prior data from NDA approved oral contraceptives. J&J should have sufficient knowledge and skill to have been able to perform the appropriate testing, development, and validation for instituting manufacturing process controls for commercial ORTHO EVRA patch production independent of input from the FDA.
J&J through the NDA process had elected to rely heavily on the performance characteristics of a few clinical lots made by Redwood and Cygnus for gaining NDA approval, while deferring until later development of the necessary manufacturing process controls until commercial launch. Commercial launch was to transition to contract manufacturer LTS. That transition to LTS as contract manufacturing did not materialize when LTS failed to be able to demonstrate bioequivalence of its manufactured lots to clinical lots made at Redwood City, CA.
During the NDA review process and prior to commercial launch; April 2002, J&J did not develop robust manufacturing release specifications and process validations necessary to ensure release of safe and effective commercial patches. As a result, the FDA provided J&J with better defined in vitro specifications which the FDA scientists derived from the NDA. The FDA's specifications include a new 24-hour in vitro testing point. As would be learned, the patch release specifications for ORTHO EVRA NDA had no relationship to actual performance of the patch when placed on a woman.
As discussed in J&J documents,(See APPENDIX 6, MANUFACTURING), the product could be released by J&J within release specifications, but had no bearing on the ability to predict the actual performance of the patch for a woman. According to testimony by Mr. Audett in 2007, J&J still had not developed in vitro to in vivo correlation for ORTHO EVRA.(APPENDIX 6 ¶ 34-52, 58-61)
#2: FAILURE TO VALIDATE THE USE OF AN EQUILIBRATION STEP
According to J&J's employee testimonies and internal documents, J&J failed to behave as a responsible manufacturer when it introduced a new "unvalidated equilibration" or "aging/curing" step into manufacturing, inserted prior to completion of in vitro testing and final release of product. J&J failed to adhere to current Good Manufacturing Practices (cGMP) when it tailed to validate and control for a critical step in the patch manufacturing process. To avoid discarding all manufactured ORTHO EVRA batches unable to meet the FDA's in vitro release specifications, early on J&J introduced the "unvalidated" equilibration step to be done prior to performing in vitro dissolution testing and packaging. The equilibration step can be variable in length for each lot and time ranged from 3 to 7 months. Use of an equilibration step prior to in vitro testing and packaging had not been described to the FDA in the NDA and would not be a manufacturing step adopted by LTS in Germany for JANSSEN-CILAG and production of EVRA. The use of an arbitrary and unvalidated aging process would also have been problematic to the FDA in terms of the Agency's concerns about changes produced by cold flow and EE drug delivery with increasing product age.
There are internal J&J documents to support that the FDA's reviewers, including Dr. Mitra, were not adequately advised of J&J's incorporation of a variable equilibration process prior to in vitro testing and product release. (APPENDIX 6 ¶ 34-52)
#3 FAILURE of J&J TO INDENTIFY AND CONTROL CRITICAL COMPONENT CHARACTERISTCIS AS PART OF EFFECTIVE PROCESS CONTROLS for COMMERCIAL ORTHO EVRA PRODUCT
As a responsible pharmaceutical manufacturer would be aware, a major basic manufacturing process control for ensuring quality and performance is adequate characterization and control of critical raw materials. J&J failed to adhere to either industry or FDA Good Manufacturing Practices (21 CFR 211) used for development of adequate written procedures and process controls necessary for manufacturing a safe and effective transdermal combination contraceptive product. J&J's ability to reference NDA approval for oral contraceptives did not reduce J&J's duty to design, test, develop and validate its transdermal patch before the launch of commercial product. As a responsible pharmaceutical manufacturer, J&J was to consistently and reliably manufacture and release a safe and effective patch intended for prevention of unintended pregnancy. Incredibly, in 2005, J&J was still in basic discussions as to how best to institute the necessary process controls, including critical component characterization, for manufacturing of ORTHO EVRA.(APPENDIX 6 ¶ 62-80)
ADDITIONAL SUPPORT FOR OPINION #3 (See APPENDIX 6 MANUFACTURING)
a. FAILURE TO CONTROL & DEVELOP RELEASE SPECIFICATIONS(APPENDIX 6 ¶ 34-52)
137. Dr. Abrams testified that when he first calculated the acceptable target range for EE for the patch as 25-75pg /mL, the value had been based on multiple pharmacokinetic (PK) studies conducted by J&J using ORTHO CYCLEN and ORTHO TRI-CYCLEN. He set the range as a "theoretical reference range" that would capture serum NGMN and EE values identified for 90% of women taking OCs. That OC drug range, when applied by J&J to a patch releasing continuous drug (EE and NGMN), would theoretically allow approximately 5% of women to have systemic serum EE values above 75pg/ml (*high serum EE values > than an OC) and 5% below 25 pg/ml (low serum EE values< OC) in terms of if she had been taking a once-a-day pill (Abrams, 6/25/07, p 85, 1, 1-12). Dr. Abrams indicated that the acceptable blood level range he developed for the patch based on OC values was "not intended to represent a therapeutic range" for a woman to receive for prevention of pregnancy, but rather provide an "acceptable target range" for drugs that are delivered by the patch (Abrams, 06/25/07 p106, 1.1-20). Dr. Abrams was unaware that anyone at J&J (before his retirement in 2004), had ever determined a therapeutic range needed for the patch (Abrams, p 107, 1. 2).
138. There are a series of internal J&J emails written in September 2001, prior to NDA approval in November 2001, about J&J's failed bioequivalence (BE) testing of LTS manufactured patch lots and the lack of "discrimination capabilities" of the dissolution test (*in vitro release test)(POE 03299371-74) The initial email was sent September 4, 2001 from Lisa Forian to Hank Avallone regarding draft meeting minutes for a meeting held August 30, 2001. She wrote (*Underlining and bold added for emphasis) (POE 03299373):
The main issues discussed were the failed bio-equivalence batches and analytical method issues. The first two Lohmann bio-equivalence batches were manufactured in West Caldwell, NJ. One failed due to high PVP swelling volume which released in high active release rate. The failure of the second lot may be due to lack of homogenization but the investigation was inconclusive. Manufacturing was then transferred to Germany. EE levels were high in the third batch. EE load was reduced by 20% for the fourth lot which is considered bio-equivalent. There are compliance concerns that the same strength product can have different formulas depending upon manufacturing location.
139. A J&J internal follow-up email was sent from Hank Avallone to Juanita Hawkins, dated September 7, 2001. The email discussed Mr. Avallone's concerns about the Draft Meeting Minutes, and the viability of EVRA and his plan to obtain additional outside consultation from ALZA (POE 03299371-74):
I have had some major concerns with regard to the viability of EVRA. The release rate specifications are very wide and it is doubtful that FDA will approve the limits established. If the CDER reviewers do approve them, it will present major problems to the operating company in the response to complaints. In this case, the major complaint, pregnancy, could have a major impact on J&J.
We have 3 failed Bio studies which is unheard of in the industry. We also have a different concentration of active in the one (4th) passing Bio study…. About one year ago PRI agreed to develop an alternate method release rate method and they have since discontinued this initiative. The release rate specification is supposed to distinguish a batch that is made properly and consistently with another that is not, yet if a batch is manufactured which is not bioequivalent, it will not distinguish it from one that is bioequivalent.
I discussed the issues and my concerns with Len Fantasia. In Lenny's former life he was very familiar with patch issues and we worked together on them. He suggested and I agree to pursue an investigation of all the in vitro results for the clinical, BE batches and current product.
I also contacted Mike Ramsay and Don Chaisson of ALZA…..They agreed that the design of the release rate test and specifications are not discriminating… Again, they suggested, as we discussed last year that we need a backup approach.
One of the things I have tried to do with J&J is to put myself in place of the FDA person reviewing this product. If I were the FDA, 1 would never accept this product with all of the outstanding questions regarding the methods and spec.
140. Dr. Polo wrote a follow-up email on September 7, 2001, but without sending a copy to Mr. Avallone (POE 03299371-74):
I do not see any new issues here. We have been telling all long that the specs should be discriminatory as a QA tool. Due to the lack of IVIVC we cannot extrapolate results to in-vivo performance. Is Hank saying now that our current specs are not discriminatory even with respect to QA?
If this is the case, is this not too late in the game? We at Regulatory, have been asking hundreds of times that the specs should be discriminatory with respect to QA. knowing that a change could be implemented without a BE study as we do not have IVIVC. If the first statement is not even true now, What do we have then?
Then we have what we always predicted that it could happen and it is a request to come back to the Phase III specs as it is in our preliminary assessment of the submission and in the assessment of the EU authorities. What I do not understand here is why now. PSGA which is the group that cannot implement the phase III specs is telling now that the current specs are not discriminatory when at the time of the filing Regulatory asked them that same question and we were told that the new specs were discriminatory.
We need to resolve this issue irrespective of FDA or any Authority, it is our duty as an organization of having one position only on this and release product with the maximum quality. Our specs are either discriminatory with respect to QA endpoints or they are not. This must be black and white to assure quality and reproducibility to our phase III Clinical data.
We cannot live in the current regulatory scenario with more ambiguity, we are too far for this, this was all discussed long time before start preparing the file.
141. Asha Ramdas (POE 02668026-7) wrote on December 17, 2001 "I specifically said that even though our drug release values for the PV(*Process Validation) and higher linespeed lots are on the low side, the plasma levels are the highest we have seen with EVRA." According to Asha Ramdas, at a meeting Mike, J&J, had responded: "further proves that the drug release values don't predict in the slightest the in vivo performance." The email continued: "However, the levels are still well within the therapeutic range and are comparable to those seen before (according to Larry Abrams)".
142. An internal J&J email sent from Hank Avallone to John Weisel dated August 30. 2001: "It's one thing to have a spec. and be unaware if there is in vitro/in vivo correlation and another to know that there is no correlation".
143. In a November 15, 2001 email, Dr. Polo discussed the history of ORTHO EVRA and the issue that was now facing J&J with commercial launch of the NDA and J&J's inability to reproduce the performance of patches used in the clinical trials:
I think FDA has taken the most conservative approach possible, but we should not lose track that it is a company problem and not an FDA one. To have only 3 lots at phase III and not being able to reproduce the phase III manufacturing conditions is a company problem.
For FDA what counts only is the conditions at which we created our phase III clinical data, the rest is out of their radar screen as we all know that we have to build our specs outside of the phase III range in order to be able to pass Process Validation. As you know we recommended to run a Bioequivalence Study Phase III lot vs. one New Specs Lot from Process Validation, in order to have the data for this potentia
l occa
sion as that will be the data that FDA could not object to. At that time the complete team decided not to pursue that path in order not to create data that could invalidate the Process Validation if at the end, the study showed no bioequivalence.
From that perspective I think that we all knew what the risks we were taking in following that path and what alternatives we had if FDA decided to stick to the tightest set of specs. I understand that sometimes it is very difficult for everyone to hear in advance the potential risks that we could have if we follow one path or the other as the interest at that time could be focus on the short future and not in the long run. When we proposed the BE study to cover for this risk, the risk of not showing BE eclipsed our proposal as the lack of BE will invalidate in the short term the Process Validation, while the gaining of a positive case of BE will only be useful in the hypothetical case that FDA will this position.
144. An April 14, 2005 email from Hank Avallone to Karen Paul documents J&J's creation of a FDA Briefing Package by Ms. Paul for a 2005 meeting with FDA and J&J's discussion of the chemistry, manufacturing and controls for production of the patch (POE 03130361):
Review (Reevalaution) of the overall scope and purpose of the presentation should be considered. If the presentation is made as is with emphasis on PK data, a legitimate conclusion could be reached by FDA to withdraw the product in the near future. It would seem that we should expect that we will not arrive at the reason for the difference in PK data between batch 1607 and the "gold standard" batch 60L005 which represents our manufacturing process. Rather than develop a risk assessment plan, perhaps a better alternative would be a plan to fix the issue. For example, a proposal to manufacture Lo Dose EVRA (similar to TriCyclen Lo Dose) could be addressed.
In addition to the CMC discussions, there should be some science/technology included. For example, we could discuss the improvements in the in-vitro methods (skin flux) and development of a better membrane to predict batch variability or evaluate changes….We comment on the limitations of in-vitro testing but offer no possible fixes.
145. On February 17, 2005, 3 years after ORTHO EVRA's launch in the United States, two complaints of pregnancy were received which were associated with the patches sold from the same lot, Lot 0403512. There were several internal emails discussing the issue (POE 04130328-9). Hank Avallone sent an email to Dr. William Randolph and Terry Dwyer, with copies to other members of the Company including to Jeff Meringer. Mr. Avallone wrote:
Terry, Judy alerted me to two complaints of pregnancy that we have received for lot 0403512. The complaint investigation in our system included reanalysis (assay only). Typically for a complaint of ineffectiveness, we would conduct a release rate test since it would pick up both total drug content and release of the drug. However, in this case our in-vitro test for picking up drug absorption is relatively meaningless. It would possibly show some effect on aging or manufacturing variability.
For some time we have been discussing the development of a release rate test that would provide in-vitro in vivo correlation (IVIVC). In situations as above, our vulnerability and overall risk to the product is significant and visible. We really can't do a decent complaint investigation without this test. We'll pick it up as a critical MAP item.
For the above batch, we need to think about doing some type of skin flux study using a Biobatch as a comparator. I would rather do that than the in vitro test in which we have a risk of failure.
146. Dr. Randolph, Senior Director Technical Development, responded to Hank Avallone and Terry Dwyer as well as other J&J employees (*Underlining added for emphasis) (POE 04130328-9):
I agree with Hank, the current methodology gives us little to work with. In addition to above situations we have no ability to determine true process control relative to effectiveness. I would support starting this work as quickly as possible as we can even if we needed to go outside. There are a number of Universities as well as outside labs that could help with this. This would be a great step forward if we could get a method that correlates to in vivo performance.
147. When asked about the issue of in vitro release specifications, Mr. Audett at his deposition conducted July 24, 2007, indicated that because of the absence of an IVIVC from J&J for commercial release, the FDA had set a very tight specification range for in vitro release testing criteria for ORTHO EVRA based on the clinical lot data information in the NDA. (NDA Supplement S001, final release specifications approved March 2004). J&J continued to conduct additional bioequivalence studies in an effort to in the future be able to return to the FDA and request to expand (widening) ORTHO EVRA's acceptable release specification range. J&J eventually returned to FDA with NDA S017, submitted January 2005, and which was approved by the FDA in July 2005. The approval of the NDA supplement allowed J&J to expand ORTHO EVRA's acceptable lower release in vitro specification] [Note:*See NRGEEP-POI-121 and NRGEEP-POI-1026 below and APPENDIX 5)] (POE 04764438)] (Audett, 07/24/07, p.375-6, 1. 9- 3)]
148. Mr. Jay Audett also testified, July 2007, that he was still unaware of the development of an in vitro test for the release of the patch that provided a meaningful correlation between product release criteria and human performance.
b. FAILURE TO VALIDATE THE NEWLY INTRODUCED EQUILIBRATION PROCESS(APPENDIX 6 ¶ 53-57)
149. There were two related emails sent in September, 2002 (POE 04104715-6) about equilibration. The first comes from Karen Paul, written as an internal email sent to Juanita Hawkins as well as others, dated September 25, 2002, following commercial launch and titled: 3-Month Protocol Hirano 2 &3 PQ and Validation Results (*Underlining added for emphasis.):
You can see from the graph above that we have three other lots where the 3- month 8-hour EE result has not significantly changed from the result reported at 4-weeks. We are conducting an investigation into these lots. In reviewing the 4-week and 3-month data currently available, the increase in release rate does not appear to be consistent….At this point there is no correlation that we can conclude.
We have captured in the MAP for the site and transition to Alza that we need to determine or have a better plan understanding as to what drives release rate and equilibration.
150. Alejandro Hojas responded to the Karen Paul on September 26, 2002:
FYI….
….Bottom line is that there is something fundamental we do not understand about the drug release and how it changes over time. There is no consistent change in result over time, also that is high variability in the data. What in our process can be causing this erratic data?
151. Karen Paul, on December 9, 2002 sent an internal email to members of J&J including Juanita Hawkins, Donna Panasewicz, Hank Avallone, and others which she titled: History 6-month Equilibration Time (POE 03169961):
Just picking your brains for a little history…
How did we arrive at the 6-month time point to perform release testing?
Answer [Panasewicz, Donna[PRDUS]]: I believe this was based on the patches ability to achieve equilibrium. Historically the time may vary from 3 to 6 months, however at six month equilibrium has been achieved. Equilibration is based upon the release rate test for all time points.
Is the test at 6-months filed anywhere?
Answer [Panasewicz, Donna[PRDUS]]: No. We have been silent on any equilibration time.
Is there a 6-months */- amount of time that is acceptable
Answer [Panasewicz, Donna[PRDUS]]: As far as I know, 6 months was established as the standard since it appears all lots at all time points have achieved equilibrium by that time.
What would preclude us from performing release rate testing prior to 6-months?
Answer [Panasewicz, Donna[PRDUS]]: From my standpoint, if we test a lot destined for normal release prior to six months and it fails I believe we may have to reject the lot. Hank would be best to address this point.
152. A May 6, 2003 email from Gopalkrishna Kurse to Juanita Hawkins suggesting that J&J produce an "Equilibration White Paper" describing the process (POE 03210533):
No matter how hard we try, the subject of Equilibration is going to come up. If we look at all the documents i.e., PQ. PV there is reference made on this docs that the Final report will be summarized at a later time point once available. So we could try our best to dodge the question but Equilibration is inherent to our process and the transdermal technology (varying degree) and sooner or later we will have to address this subject and thus why not today. Of course we will only bring this White Paper up if we see that it is practically impossible to have a dialogue without bringing this subject. We could write a White Paper on Equilibration and address this subject using science….Manati will benefit from this White paper.
This document should cover the chain of events leading to this 6* month hold/equilibration time.
153. Juanita Hawkins' reply to Gopalkrishna Kurse (POE 03210533):
Good challenge. Where I am coming from is not that I don't want it discussed, but just as Duragesic, from a compliance perspective it would be a risk if we chose to vary the equilibration time. The issue is that if the equilibration time varies from 3-7 months, then how do we substantiate we have a validated, repeatable process? We would argue that of course and say that we are working to understand the science, but as in Duragesic we never came to the real understanding that we could test at a decision time and stand by it….the lot is good or the lot is poor. We look at the data, look at the curve profile and make a decision. Not all investigators are ready to deal with science. Additionally, if we are going to discuss equilibration we would have to be prepared to answer why the equilibration time is lengthening and I haven't heard that we have done a complete investigation of root cause, documented it and passed a strong challenge.
c. FAILURE TO INDENTIFY AND CONTROL CRITICAL COMPONENT CHARACTERISTICS AS PART OF EFFECTIVE PROCESS CONTROLS
(APPENDIX 6 ¶ 62-83)
154. (Exhibit 757)- February 17, 2005 chain email from Dr. Randolph to Judy Dowling, with CC to engineer Mr. Jeff Meringer, said:
I agree here with Hank, the current methodology gives us very little to work with. In addition to above situations we have no ability to determine true process control relative to effectiveness.
1. &J's LACK AND DELAY IN ADEQUATE CHARACTERIZATION OF CRITICAL COMPONENT-PVP(APPENDIX 6 ¶ 62-83)
155. There appears to have been a J&J and ALZA meeting during the third quarter of 2003, after market launch, to better understand the "PVP issue" (POE 00146196-7). Both Redwood City(RWC) and LTS had received two batches of PVP material within specification yet out of trend in terms of particle size and bulk density. LTS ran approximately 30 batches all of which tested too high for release for both EE and NGMN. RWC produced 63 high EE out of trend batches using two different PVP lots from BASF(one batch yielded 47 batches, the second 16 batches). All 63 batches called fast equilibrating (in 2-3 months vs. the usual 6-7 months), and were not released commercially. Since there is no IV/IVC test method available for this product they were hand pouched. In addition, there was one patch lot that had a one-month time point taken when being kept at 40 degrees to attempt to help accelerate equilibration. This batch trended out of specification. The root cause for being out of trend was identified as the source of PVP batches used. The same PVP source batches had produced the same types of performance issues when used for patch manufacturing either in California or in Germany by LTS. The final points raised during the meeting were as follows:
• At the current time we do not have an in-vitro/in-vivo correlation test; therefore we cannot predict from this data how the product will react in the body.
• We do not have a good in-depth study of PVP or other major raw material influences and specification ranges for the product; as a result we operate within the approved specification ranges for the product; any out of trend data is investigated as demonstrated by this study. This process will remain in place until we have better scientific information and some PK or bioequivalence data.
156. There is a March 23, 2004 summary of J&J's IDC Meeting Minutes (POE 00732011-2):
Factors identified as most likely to effect ethinyl estradiol (EE) release were considered to be:
• Polyvinylpyrrolidone (PVP physical properties)
• PVP dispersion
• Isopropyl alcohol amount (IPA)
• PVP amount
• Lauryl Lactate composition
• EE loading
157. On August 05, 2005, Jay Audett sent an email to Michael Jackson regarding the need to control raw materials (POE 02705493):
I want to update you on the discussion I had with Gary Neill today regarding EVRA. He is a strong proponent for the development of a less variable formulation with reduced estrogenicity and understood the importance of gaining control over key raw materials. I made it clear to him that it will take at least a year to sort through raw material factors because of technical challenges at BASF, and that we needed to work more effectively with PSGA to make sure that both groups give this appropriate prioritization. He offered to provide assistance, if required. I responded that I would sort through this with Bill Randolph. To make this happen, PSGA needs to provide more than a fraction of a head count. I have since spoken with Bill at PSGA (who is not very happy with me), and he acknowledged the need to re-center delivery and pledged to work with us to develop a plan, assigning a full head count to bring this to conclusion.
158. On November 8, 2005 ALZA made a presentation regarding ORTHO EVRA Crospovidone (PVP) and Influence on Patch Performance. The scheduled agenda was to discuss "Why Crospovidine was Critical"(POE 01998225). The program began with a review of PK studies that established the EE dose, reviewed skin flux studies relative to clinical outcomes; Crospovidone material review; and lauryl lactate drying studies in Manati. The operational plan at ALZA in 2005 was to develop a new low EE version patch as soon as possible and demonstrate bioequivalence using a new adhesive formulation. ALZA Development Group's goal was to complete Crospovidone manufacture optimization and characterization for ORTHO EVRA to support commercial product and develop a low EE product.
159. The "Equilibration rate" of drug appeared to be dependent on Crospovidone surface area (particle size distribution), quality of dispersion, and possibly oven residence time. "Crospovidone is the most critical raw material component". Potential critical attributes of PVP include surface area, particle size, swell volume, pore size and dispersion quality.
160. In the history of ORTHO EVRA's Development Changes, it had been found that the amount of EE was to be increased for phase 1 to achieve higher in vivo delivery – the in vitro skin flux was predictive for the increase. During the phase 2 study batches, the mix utilized higher shear: in vitro skin flux was used to be able to predict higher EE delivery and lower NGMN delivery. It was hypothesized that the isopropyl alcohol (IPA) content in the mix needed to be higher because of insufficient Crospovidone softening that slowed drug partitioning while mixing, unless higher shear force was employed. During the production of the phase 3 clinical lots, with the increased IPA levels, there was a switch to less complex lauryl lactate mixture, and use of TPM (higher shear than for phase 1). The development studies prior to clinical lot production showed that variations in mix time did not impact in vitro skin flux. IPA increase was correlated with higher EE skin flux. A switch to less complex lauryl lactate countered the EE flux increase and enabled EE delivery identical to phase 2.
161. For the summary of changes to the "Commercial Lots," the introduction of a faster line speed (change from 1 meter/min to 1.7 meters/min) impacted the equilibration and caused higher QC release. Crospovidone (PVP) raw material changes made prior to commercial launch with different PVP particle size distribution was associated with out-of-trend lots, but provided a clue as to J&J's need to tighten the Crospovidone specifications.
2. Path To Commercialization (ALZA SUMMARY November 8, 2005):
TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE
• Prior to validation and commercialization, Crospovidone vendor could not reproduce phase III quality material
• Is the current Coviplas material identical to the phase 3 Crospovidone?
162. There are factors that have been identified that can influence EE delivery: Alcohol content in mix- evidence in phase 2. phase 3 and LTS mix; the lauryl lactate composition- with a change from C12, C14, C16 blend alcohols to all C12 composition alcohol-development studies prior to phase 3 production and Manati process modification; Crospovidone particle size distribution (or other material attribute)- still to be determined; BASF's(PVP supplier's) establishment of better process control at its MP plant. (MP: One of two BASF supplying facilities for PVP )
163. The conclusion of ALZA's 2005 PVP presentation, made 3 years after product launch and commercial marketing: there must be careful evaluation of any change made in materials or process as they moving forward; balance QC drug release test requirements with EE delivery; and utilize Flux HTS capability at Transform Pharmaceuticals. This is the type of basic GMP conclusions regarding the importance to control raw material that would be made during development and testing of a product prior to NDA approval. It also reflects the FDA's allowance of J&J to bridge prior ORTHO-CYCLEN and ORTHO-TRICYCLEN ORAL Contraceptive NDAs for design and development of a transdermal contraceptive patch.
3. COMPONENT- LAURYL LACTATE
164. Laury lactate is a patch penetration enhancer. DECEMBER 13. 1999 NDA MAA TEAM MEETING MINUTES sent out by Valerie Donnelly from December 15, 1999. (POE00347151-54) (Exhibit 688) (APPENDIX 6 ¶ 3.8.68.71,77,143,150,173,203,208,214,226,254,27-77, 340-1) stated:
NDA ITEM 5, Nonclinical Pharmacology/Toxicology/ADME
As decided at the Global Product Team meeting on December 13, 1999, the three lauryl lactate study reports will not be added to the NDA. Therefore, this item can be considered final. A second draft publishing to verify revisions to the references will occur in late January 2000.
165. Dr. Mitra had questioned J&J's evaluation of lauryl lactate (LL) in his initial review of the NDA. Dr. Mitra acknowledged that this excipient, which is the penetration or flux enhancer for EE, could affect EE delivery. Ms. Paul, J&J, told Dr. Mitra that the LL was within specification for both the clinical development and commercial batches.
166. According to the NDA Review by Dr. Chatterjee, on September 28,2001, and prior to NDA approval, J&J notified the FDA (upon inquiry by the FDA) that the formulation it had used in study NRGEEP-CONT-001 contained "5.5% lauryl lactate" and would be different from the final "to-be-marketed" commercial formulation that would contain a decrease to 3.78% lauryl lactate.
167. In J&J's summary of All CMC Changes (POE 04764435), it lists this NDA supplement (S007) approval. J&J indicates that a new analytical method was instituted for determining the identify and chromatographic purity for lauryl lactate, and to change assay specification. This was a 30-day CBE-30. Factors most likely to effect ethinyl estradiol (EE) release: PVP physical properties; PVP dispersion; IPA amount; PVP amount; lauryl lactate composition; EE loading.
4. DESIGN OF EXPERIMENT(APPENDIX 6 ¶ 303)
168. The Design of Experiment (DOE) – examined 5 factors: PVP particle size, PVP swell volume, PVP dispersion, IPA amount, PVP amount result in 16 trials for half fraction experimental design. It resulted in showing all main effects and 2 defining factor interactions. The DOE proposed to include two analytical techniques to evaluate each experimental lot's performance- in vitro testing: current quality control test for drug release and the addition of skin flux testing- a good estimate of in vivo performance but not to the level of an in vitro/in vivo correlation (IVIVC). Skin flux testing was done in the early ORTHO EVRA development and had success in providing inference to true in vivo performance.(POE 00127209). ALZA had an ongoing collaboration with TransForm Pharmaceutical to develop high throughput screening (HTS) tools for transdermal evaluation.
OPINION #4:
169. J & J failed to update its product labeling and promotions with warnings in a timely manner regarding the increased post-market reporting of unintended pregnancy and adverse events including myocardial infarction, VTEs and death. (See OPINION #1 as well as APPENDIX 6 ¶ 511-522, 536-588 ). J&J heavily promoted its transdermal patch to women as a simple, convenient, once-weekly dosing for prevention of unintended pregnancy without having obtained scientific support to justify a new increased compliance claim within its NDA, a point discussed with J&J by FDA during the End of Phase 2 Meeting in 1997 (POE 00013856). Such conduct showed a disregard for J&J's duty to truthfully sell a safe and effective product to healthy women utilizing adequate promotions and warnings so that both physicians and women can make informed decisions as to risks and benefits of elective use of ORTHO EVRA versus another contraceptive method for prevention of unintended pregnancy (APPENDIX 6 Bakhru, August 2006 ¶ 526-528; 117, Thurman 2007 ¶ 546-8).
ADDITIONAL SUPPORT FOR OPINION #4
a. FDA's REVIEW OF FILED ADVERSE EXPERIENCE REPORTS FOR ORTHO EVRA (APPENDIX 6 ¶536-538)
170. According to the FDA's January 2007 Review of ORTHO EVRA AERS. January 16. 2003 was the first FDA DDRE Safety Update Report provided in response to an earlier request for a consult from the Division of Reproductive and Urological Products (DRUP). Consult: PID #D020374 (FDACDER 001193).
b. J&J's POST-MARKETING SAFETY UPDATE- YEAR 2003(APPENDIX 6 ¶ 270-273)
171. There were a series of slides for a presentation by P. Caubel, M. Cobb, and D. Fife(POE 00216399-412) (Exhibit P-61) dated November 6, 2003 (Data from adverse event reports April 2002-October 2003). From April 2002 to October 2003, EVRA had 909,394 women years (WY) of exposure, and TRI-CYCLEN had 4,028,541 WY. In terms of differences in effectiveness for prevention of unintended pregnancies, there have been 433 pregnancies reported for EVRA ( 476.14X 106) ( pregnancies/exposure) and 86 (21.35 X 10-6) for TRI-CYCLEN ( EVRA with a 22X increase in reporting compared to TRI-CYCLEN); 18 CVAs for EVRA (19.7 x 10-6) vs 8 CVAs (1.9 x 10-6) for TRI-CYCLEN( EVRA with a 10X increase in reporting compared to TRI-CYCLEN) . For the 18 CVAs reported for EVRA, 13 (72%) were medically confirmed, in comparison to the 8 CVAs for TRI-Cyclen, 3 (38%) medically confirmed. In eight of the 13 medically confirmed CVAs with EVRA, the women each had 1 risk factor. Two of the 3 medically confirmed CVAs with ORTHO TRI-CYCLEN each had one risk factor.
172. By November 2003, there had been 4 women with MIs reported for EVRA and 1 for ORTHO-TRI-CYCLEN. Twenty-two pulmonary embolisms (PEs) were reported for women with ORTHO EVRA and 9 for TRI-CYCLEN. In terms of PE/exposure, there were 24.1 x 106 EVRA and 2.23 x 10-6 ORTHO TRI-CYCLEN (an increase by approximately II times for ORTHO EVRA). "Other reports" of thromboembolic cases, included 35 for EVRA and 10 for ORTHO TRI-CYCLEN (an increase by approximately 4 times for ORTHO EVRA versus ORTHO TRI-CYCLEN). Regarding the "Other reports" of thromboembolic exposure, there were 38.5 x 10-6 for EVRA and 2.5 X 10-6 for ORTHO TRI-CYCLEN (an increase of approximately 15 times for ORTHO EVRA compared to ORTHO TRI-CYCLEN).
173. For the Spontaneously reported cases, of the total that were deemed evaluable, 322 were for EVRA and 77 for TRI-CYCLEN as of August 2003. Regarding the increased reporting of pregnancies, there was a recommendation in 2003 that COMPREHENSION STUDIES, including both label comprehension and user comprehension studies, be conducted.
SAFETY CONCLUSIONS (1)
• Is there a PK trend of concern?
-Commercial batches with higher EE delivery rate than clinical batches?
-Equivalence between LTS and Redwood City.
• Is Labeling Still Valid?
-20 mcg based on IV study: still appropriate?
-Table of PK data: to be revisited based on NED-1 study results?
-Comparability language with OCs to be updated?
-Communication with Health Authorities?
SAFETY CONCLUSIONS (2)
• Is there an effectiveness issue?
-Is appropriate use of the product well communicated and well understood by patients and HCP?
-Access to replacement patches to be investigated and harmonized?
• Is the number of thromboses (arterial and venous) "out of range" compared to other hormonal contraceptives?
-Can it be linked to a higher exposure to EE (plasma levels and pattern of administration)?
-Are some additional investigations to be proactively undertaken?
TEAM RECOMMENDATIONS
• Development and Implementation of Risk Management Plan which may include:
-Continue monitoring/heightened surveillance and assessment of case reports
-User comprehension study in US to evaluate pregnancies
-Epi study in the EU
-Epi study in US to assess risk of CVAs/PE/DVTs
• Audit lab regarding EE assay methodology
• Evaluation of AUC of commercial product
• Develop & Implement Proactive Regulatory Strategy for NED-I study
• Development of IVIVC method
c. J&J's EXPERTS PANEL MEETING RECOMMENDATIONS FOR PROSPECTIVE STUDY- FEBRUARY 8,2004
174. On February 8, 2004 a "global panel of renowned clinicians (key opinion leaders "KOL") were invited to J&J's initial post-marketing review of ORTHO EVRA. The agenda included 5 presentations including I) clinical overview of ORTHO EVRA; 2) unique marketing approach and history; 3) review of pharmacokinetics and pharmacodynamics: 4) coagulation review and data on the "starter" phenomenon; and 5) spontaneous reports of adverse events along with comparator Yasmin and FOI data.(POE 00127641-661)(Exh P-138). The KOLs were asked by J&J to consider: did the pharmacokinetic trends in EE exposure seen in the manufactured lots compared to the clinical lots explain the "post-marketing safety experience." With respect to the reported pregnancy events, did they suggest a pharmacologic issue or a user issue with the patch? What is the role of the unprecedented marketing efforts in identifying and reporting adverse events (AEs)? Was the clinical lot too estrogenic?
175. Within 20 weeks of the direct to consumer marketing (DTC) campaign, ORTHO EVRA had become the second most popular brand of hormonal contraception, and had gained the second highest share of the contraceptive market. The targeted population was women aged 18 through 34 years. The expert panelists recommended that J&J conduct a possible prospective study to evaluate the validity of the spontaneous reports being received. The expert panel members were not provided the FDA's risk calculation of VTE 118/100,000 WY, but rather J&J's 59/100,000 WY as issued by J&J in its Investigator's Brochure (POE 00127649).
176. The KOLs discussed the clinical lot and the increased effects of estrogen (estrogenicity). In a clinical trial comparing ORTHO EVRA versus Mercilon (desogestrel) and Triphasil (levonorogestrel) regarding hemostasis values in 103 women, (NRGEEP-CONT-006), SHBG (sex hormone binding globulin) levels were higher in ORTHO EVRA groups than with either OC at Cycle 1 and Cycle 6. Therefore, ORTHO EVRA had a stronger estrogen effect than OCs. Regarding C-reactive protein levels, which has been identified as a strong risk factor for stroke, they were increased for all treatment groups with differences between the three groups not significant. Fibrinogen levels, considered a risk marker for arterial events, were increased for all treatment groups, but again with no significant differences between groups. There was a trend seen for ORTHO EVRA to increase prothrombin fragment 1*2 (F *2) – a marker of coagulation activity – again more than the OCs.
177. Regarding the prothrombin fragments 1*2, according to Dr. Friedman, J&J internally was aware that the absolute values for F1*2 values were not statistically significant; however, examined as percent values, there was statistical significance in terms of ORTHO EVRA to comparator OCs. Measurement of circulating levels of prothrombin fragment 1*2 (F *2), according to the American Heart Association, is considered a specific marker of thrombin generation in vivo, a diagnostic tool in the evaluation of the hypercoagulable state. [J.A.Paramo, et al. Prothrombin Fragment 1*2 Is associated with Carotid Intima-Media Thickness in Subjects Free of Clinical Cardiovascular Disease, Journal of American Heart Association, 2004;35;1085-1089] (APPENDIX 6 ¶ 30, 109, 178, 198, 222-225).
178. The results of CONT-006 combined with the 1998 reports of increased VTE risk in the clinical trials ( 18/100,000 WY), are both safety signals that J&J had regarding the increased risk of cloning with ORTHO EVRA when it was compared to 35 mcg OCs. J&J possessed both of these coagulation safety signals as well as NRGEEP-PHI-017 prior to its pre-NDA negotiations with the FDA in July 1999, yet the data supporting increased risk to OCs was not discussed with the FDA and was not fully disclosed to the FDA in the NDA. Such information should have been used by a responsible manufacturer before it began commercial release of a product like ORTHO EVRA, [NGMN * EE] patch, to increase its warnings to physicians and women regarding increased venous thrombotic event risk of the patch compared to OCs. Prior to NDA approval, J&J was aware that the risk of VTE was increased when ORTHO EVRA was compared to 35 mcg EE OCs, each with a different progestin, desogestrel and levonorgestrel. J&J also should have been aware of its duty to aggressively monitor public safety and to quickly and fully communicate the risk to physicians and patients prior to commercial release of ORTHO EVRA and with post-approval updates to the product information and labels.
179. Despite J&J's knowledge of the multiple safety signals of increased risk for thrombotic events compared to OCs, J&J persisted in continuing to inform its clinical investigators and experts that there was no difference in effects on two fibrinolysis activity markers, fibrin degradation products d-dimer (FDP d-d- or plasmin alpha-2-antiplasmin (PAP). However, the effect of ORTHO EVRA and its 17-d-NGMN on APC-resistance test was shown to be more similar in effects to a known 3rd generation progestin, desogestrel than to the parent progestin norgestimate.
180. The experts were informed that ORTHO EVRA had a lower effect detected on Protein S (a coagulation inhibitor) when compared to OCs. The change from baseline in prothrombin time was similar for ORTHO EVRA and OCs for all time points. (POE 00127652).
181. The February 2004 Executive Summary discussed a "Starter/Phenomenon" with an elevated risk of VTE seen in starters of OCs and "HRT" decreasing from 10/10,000 to 2-4/10,000 for ongoing OC/HRT use. "In other words, there appears to be an excess number of VTE among new starters of hormonal contraceptives that stabilizes shortly after treatment initiation." Regarding "estrogenicity," ORTHO EVRA appeared to be an "estrogen dominant" hormonal contraceptive as determined by its "effects on lipids and the predominance of estrogen-related side effects." Pharmacokinetic studies had shown that the mean area under the curve (AUC) levels of EE for commercial lots are greater than those for the clinical trial lots; however. there was no difference in NGMN levels. The reason for this difference in EE was not understood. The variability analysis demonstrated "greater between subject variability than lot-to-lot and study-to-study variability". Of importance was the finding that the "mean exposure from the commercial lots is 122% of the mean exposure from the clinical lot – demonstrating an excess EE of 22% from two recently completed PK studies." "This finding raises the question of whether the two lots (clinical vs. commercial) are bioequivalent." (POE 00127653-4).
182. Regarding the findings in PRI-NED-1, the summary states: A PK-PD trial found the EE A UC exposure was greater for ORTHO EVRA than for CILEST (NED-I). In addition, ORTHO EVRA exhibited similar/albeit somewhat higher- plasma concentration of SHBG compared to Cilest. Currently, there are not sufficient data to draw conclusions regarding the risk of VTE associated with use of ORTHO EVRA, "in part because the rates remain so low. There is a need to establish a clear database regarding the risks of VTE with use of transdermal contraceptive patch, and this has been undertake." (POE 00127654).(APPENDIX 6 ¶ 283-288)
183. The February 8, 2004 recommendations from a consensus of the Expert panelists and KOLs was that:
ORTHO EVRA appears to be an estrogen-dominant method; as such, they agreed that decreasing the estrogen dose would negate the need for a large-scale epidemiological study. If the dose is not reduced, the panel recommended a large prospective case-control epidemiology study looking at the risk of VTE and CVA, using a 35 mcg OC as a comparator. The group generally agreed this would be a very difficult and large undertaking.
The Company was congratulated on taking a pro-active course prior to any required FDA interventio
n, bu
t concurred with the plan to rapidly brief FDA of these findings." (POE 00127655-6).
d. IDC MEETING FEBRUARY 2004- REQUEST FOR FUNDING(APPENDIX 6 ¶ 290-293)
184. There are meeting minutes (POE 00520104-5)(Exhibit P-130) of the IDC (Integrated Development Committee) held on February 17, 2004. The Team Reviews include "EVRA: Funding for Benefit/Risk Program" with presenters Angie Guisti, Bob Kale, Naresh Punwani, and Patrick Caubel. The issues discussed were that the EVRA team , request to the IDC to allocate funds ($11,01MM in 2004 and $5.01MM in 2005) and resources to conduct an Evra Risk Assessment Plan, as proposed at the November 6, 2003 IDC meeting and the December 17. 2003 NPDC meeting. The risk assessment plan would include: continuation of heightened vigilance and analysis of SAEs; conduct PK studies to evaluate the commercial lots; conduct PK studies to qualify the Manati manufacturing sites; and assessment of the feasibility of a US epidemiology study.
185. At the conclusion, the IDC deferred approval of the requested funds. The IDC emphasized the "need for a sense of urgency in bring the product 'under control"'. (POE 00529105).
e. IDC/PPM/PSGA MEETING MARCH 2004-REQUEST FOR FUNDING (APPENDIX 6 ¶ 294-308)
186. There is a slide presentation for an EVRA/ORTHO EVRA IDC/PPM/PSGA March 23, 2004 program, again with a review of a request for funding (POE 00127189- 221). With presenters (See above February and March, 2004 meetings) again Angie Guisti (Introduction), Bob Kale (Resources), Naresh Punwani (PK program), and Patrick Caubel (Benefit/Risk Programs). A decision was needed from the IDC/PPRM/PSGA for an allocation of funding and resources to conduct an EVRA risk assessment as presented to the IDC (November 23, 2003). NPDC (December 17, 2003), IDC (February 17, 2004). For PK, the ethinyl estradiol should be further investigated. This is due to data from a limited number of lots (2 commercial/2 validation lots out of 150 commercial lots) suggesting increased EE exposure compared with lots used in the clinical program. Seventy-six lots of product were manufactured with PVP that caused high in vitro EE drug release- 61Out of Trend ORTHO EVRA Lots and 15 Out of Specification EVRA Lots. There were high numbers of ORTHO EVRA spontaneous case reports received with a large volume of thrombosis-related events and pregnancies when ORTHO EVRA was compared to J&J's internal OC comparator (TRI-CYCLEN) marketed by J&J since 1992. The cause was unclear to J&J (POE 00127192). For the Out of Trend U.S. commercial lots, the impact on manufacturing process steps and raw materials on EE absorption and in vivo performance was unclear to J&J. There was now a high incidence, (safety signals) of thromboembolic events and increased reports of pregnancy.
f. SIX LOTS WITH INCREASED REPORTING OF SPONTANEOUS THROMBOTIC EVENTS (APPENDIX 6 ¶ 317)
187. Karen Paul sends an email dated April 15, 2004, regarding CMC Piece to "Len, Terry, Juanita and Hank" (POE 03130415-8):
For the introduction letter…Results of NED-I Study…We are taking the following actions
For the overview-
All lots released for distribution have met all specifications as approved in our NDA. There were a total of 6 specific lot numbers associated with the spontaneous post-marketing reports of VTEs. The lots were reviewed and there were no deviations associated with them that would affect product quality, and the assay and impurity results from testing retained samples are within the approved product specifications. No trends were noted.
For risk assessment-
In response to the present PK data, a CMC investigation was initiated. The investigation has focused on three key areas: an evaluation of the lots associated with spontaneous post-marketing reports of VTEs, a review of the manufacturing process and development of a DOE to further characterize the impact of key raw materials and processing variables on product performance.
Data Assessment-
Lot numbers were provided for seven of the post-marketing spontaneous reports regarding VTEs. Of those, six were valid lot numbers. The batch records for those lots have been reviewed, and there were no deviations associated with the lots that would adversely affect product quality, the assay and impurity results….
All lots released to distribution have been within specifications as approved in our NDA. Post-approval changes have been the subject of either a supplement (PAS or CDBE-30) to our approved NDA or included in the Annual Report (most recent filed January 2004)
Risk Assessment Intervention
Based upon the retrospective review of the product and manufacturing process, a DOE is being developed to further characterize the impact of key raw materials and process attributes on product performance. [Paul, Karen[PSGA]]. In addition, we are evaluating the possibility of revising the formulation to reduce the EE delivery to remain bioequivalent to the current commercial product toward the lower side and more centered on the clinical batch.
g. SPONTANEOUS REPORTING OF PREGNANCY AND VTES – APRIL 13, 2004 (APPENDIX 6 ¶ 318)
188. Key observations included a significant volume of spontaneous reports on ORTHO EVRA, including pregnancy and DVT (POE 00722686-695)(Exhibit 114, Stepic). The PK data on 2 commercial lots suggest increased EE exposure vs. clinical lots- Product is within limits of approved label. Sixty one lots of US commercial patches are out of trend in terms of EE in vitro release. Risk assessment program support was requested of NPDC- April 14, 2004.
189. The main issues driving exploration for ORTHO EVRA were spontaneously reported Adverse Events (i.e. pregnancy, DVT) and the differences in systemic EE release of commercial vs. clinical lots from Redwood City (and vs. LTS). The spontaneous reports of pregnancy were reviewed by PRD, then by outside experts. All AEs were numerically higher than ORTHO TRI-CYCLEN, but attributed to be likely due to significantly higher consumer & physician marketing. Reportedly, the outside expert did advise J&J to look further at DVT. A lull AE review was conducted internally, with a focus on DVT. An Expert Panel was convened on February 2004. The PK observations were that ethinyl estradiol, in particular, should be further investigated. All commercial lots released had been able to meet manufacturing release specifications. On March 24, 2004, CDT sought support from IDC. Benefit Risk Assessment by CDT to IDC was that the current risk/benefit balance of ORTHO EVRA showed safety profile and effectiveness rate within the limits of approved label and consistent with a newly approved hormone contraceptive. CDT presented the same to NPDC (4/14/04). IDC did not approve US Epidemiology study funds at this time.
190. A pulmonary embolism death was reported in NYC last week of a young woman believed to be on ORTHO EVRA (details unconfirmed).(POE 00722690). Additional PK and manufacturing work was proposed to help further assess commercial performance vs. clinical lots and labels. CM&C was to evaluate multiple process and raw material impact on in vitro performance and delivered EE dose. If needed, the "manufacturing process will be changed" (POE 00722691).
h. J&J's EPIDEMIOLOGY STUDIES (APPENDIX 6 ¶502, 503-508, 543-545)
191. According to the FDA's Review of ORTHO EVRA AERS on "January 10, 2006, the epidemiologic study by SS. Jick et al, was now available online at Contraception" (FDACDER 001194). This article supported that there was a lack of increased risk for nonfatal VTE associated with ORTHO EVRA. When published, SS. Jick et al. was titled Risk of nonfatal venous thromboembolism in women using contraceptive transdermal patch and oral contraceptive containing norgestimate and 35 ug of ethinyl estradiol; Contraception 2006;73(3):223-228.
192. The second epidemiology study, also funded by J& J, was J. Alexander Cole, et al. (Venous thromboembolism, myocardial infarction, and stroke among transdermal contraceptive system users, Obstetrics and Gynecology, Vol. 109, No.2, Part 1, February 2007). The 13 study attempted to estimate the incidence of venous thromboembolism (VTE), acute myocardial infarction (AMI), and ischemic stroke associated with ORTHO EVRA users compared to users of approved norgestimate-containing OCs. According to the authors, it was given a contract from J &J which granted it oversight of the study conduct, report and interpretation, and final wording. As the study progressed, both J & J and the FDA were to receive interim reports. The primary source of data was a database of eligibility, pharmacy claims, and medical claims from UnitedHealthcare, a major national health insurer. The claims data was supplemented using pharmacy dispensing records and National Death Index Plus through year 2003. The authors identified women exposed either to ORTHO EVRA or norgestimate-containing OCs from April 2002 through December 2004. The norgestimate-containing oral contraceptives were to have 35 mcg ethinyl estradiol (EE), the approved norgestimate COCs.
193. There were 49,048 woman-years (WY) of ORTHO EVRA exposure from April 2002 (USA product launch) to December 2004. This was compared to 202, 344 woman-years of norgestimate OC exposure with 35 mcg EE over the same time period, producing roughly a ratio of 1:4 dispensing of ORTHO EVRA to norgestimate-OC woman-years of exposure. There was a more than two-fold increase identified in VTE rates reported among ORTHO EVRA users when compared to norgestimate-containing OCs. Women using the patch had average levels of circulating estrogen measured as 60% higher than women using oral norgestimate contraceptives pills, although the peak level (C Max) would be lower for the patch user.
194. The entire Cole ( 2007) study population included 340,377 women who received at least one dispensing of ORTHO EVRA (98,790) or one norgestimate-containing OC (256,981) between April 1 2002 and December 31, 2004. The median age of women was 25 years. For VTE, 33% of cases and matched controls were between the ages of 15 and 29 years, 57% of cases (55% of controls) were between the ages of 30 and 39 years, and 10% of cases (12% of controls) were between the ages of 40 and 44 years. For new initiators, the odds ratio (O.R) for VTE associated with use of ORTHO EVRA was 2.2 (95% CI 0.8-6.1), for women who were not new users, the O.R. for VTE with ORTHO EVRA compared to norgestimate-OC was 2.1 (95% CI 0.8-5.7). Among ORTHO EVRA exposed cases, 50% had 90 days or less exposure, compared to 48% controls. For norgestimate-OC exposed cases, 66% had 90 days or less exposure, compared to 41% controls.
Table 1. Incidence Rates and Rate Ratios
ORTHO EVRA … NGM-OC
TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE
Table 2. Odds Ratios, Nested Case-Control Analysis
ORTHO EVRA … NGM-OC Users
TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE
i. FDA's ANALYSIS OF AERS
FDA FOUR MONTH SAFETY UPDATE REPORT- MARCH 1, 2006- (See APPENDIX 6¶500-510)
195. According to the FDA's January 2007 Review of ORTHO EVRA AERS, this was the fifth DDRE Safety Update Report provided in response to a request for a consultation by the Division of Reproductive and Urological Products. Consult: PID #D060097 (FDACDER 001193).
196. Adrienne Rothstein, Pharm. D and Rita Ouelle-Hellstrom, Ph.D., M.P.H. DDRE, sent a March 1, 2006 Four Month Safety Update to Dr. Shames, Director, DRUP for ORTHO EVRA, Yasmin, NuvaRing, and Alesse. The report went through Mark Avigan, M.D, C.M, Director DDRE. The reactions being reviewed in the FDA's AER database were Serious Thromboembolic and Thrombotic Events, including Pulmonary Embolism, Deep Vein Thromboembolism, Cerebrovascular Events and Death. (FDA CDER 000870-91). This report was made to update previous consults of 02/14/2005 and 11/08/2005.
197. On February 14, 2005, the Division of Drug Risk Evaluation (DDRE) provided a summary of serious thromboembolic and thrombotic (TT) adverse events in women using Ortho Evra, compared to similar events with Yasmin, Alesse, Mircette, Ortho Tri-Cyclen Lo, OrthoCept, NuvaRing, Norplant and Mirena. In addition to Ortho Evra, reporting rates were calculated for Alesse, Mircette, Ortho Tri-Cyclen Lo, and NuvaRing.
198. Due to a continued need to evaluate reports of serious TT events for this drug class, the DRUP and DDRE met on July 19, 2005 to discuss periodic updates. All domestic and foreign AER reports were reviewed through December 31, 2005. In the summary of the data, the age range for TT reports for Ortho Evra ranged from [12-53 years, with the average age 29.2, and median age 28 years. The FDA had received "336 reports" of a woman with a TT cases. Of those reports, 42.8% PE, 39.4% DVT, 21.1% CVE (cerebrovascular event) cases, 7% other cases and 2.8% MIs. The average duration of use was 6.1 months (range 0.03-40 mos), median duration 3.0 months. The average weight 73.3 kg (range 45-159kg) and median weight 70.9 kg. Average # of TT risk factors 1.0 (range 0-4), median # of risk factors 1.0.
Women Younger Than 30
199. Compared to Yasmin, Ortho Evra had higher reporting rates for all TT events per 100,000 person-years (9.9 vs. 3.7), PE (4.0 vs. 2.1), and DVT (4.2 vs 1.8). Ortho Evra reporting rates are four times greater for CVE (2.2 vs 0.5), OTE (0.5 vs 0.1), MI (0.3 vs. 0.0) and for deaths for all causes (0.7 vs. 0.1) when compared to Yasmin. There are no MIs for Yasmin among women less than 30 years of age.
200. When compared to NuvaRing, Ortho Evra has similar reporting rate for PE (4.0 vs. 4.5) in this age group and for CVE(2.2 vs 2.1). Reporting rates for all other TT events and deaths from all causes were higher for Ortho Evra: DVT (4.2 vs. 1.2), OTE (other thrombotic event) (0.5 vs 0.0), Ml (0.3 vs 0.0) and all deaths (0.7 vs 0.4).
201. When compared to Alesse for the first five years of marketing, Ortho Evra has over tenfold greater difference in reporting rates for all thromboembolic events and all cause mortality either combined (9.9 vs 0.5) or analyzed separately.
202. When reviewing the adverse event reports retrieved from AERS, of particular concern to the FDA was the number of CVEs and MIs observed with Ortho Evra compared to the other combined hormonal contraceptives. "When the U.S. reports of CVE ('cerebrovascular event') are examined in detail, Ortho Evra is the only product with reports of fatality. The patient ages for fatality are striking, with 4 of 5 patients aged 17-25 year old; the remaining patient was 37 years old at the time of death."
Women Older Than 30
203. Compared to Yasmin, Ortho Evra has slightly higher or similar reporting rates for all TT events per 100.000 person-years. Compared to NuvaRing, Ortho Evra has a higher reporting rate for CVE (3.7 vs 1.6) and PE (8.3 vs 6.0). Ortho Evra has lower reporting rates for DVT (6.2 vs 7.6) and OTE (0.5 vs 1.6), and the same reporting rate for MI (0.5 vs 0.5). NuvaRing has no reported deaths among women 30 years of age or older compared to 5 for Ortho Evra. When compared to Alesse and the first five years of marketing, Ortho Evra has significantly greater reporting rates for all thromboembolic events combined (16.1 vs 1.4) or analyzed separately.
FDA'S EXECUTIVE SUMMARY (Ortho Evra only):
The reporting rate analysis continues to show that Ortho Evra has slightly higher reporting rates for serious TT events than Yasmin and NuvaRing, and significantly higher reporting rates when compared to the first five years of marketing for Alesse. The increase remains more readily apparent for MIs and all cause mortality. The differences in reporting rates are particularly noticeable for women younger than 30 years of age. Publicity and litigation appear to have negatively impacted the total number of prescriptions dispensed in 2005 and stimulated an increase in the number of adverse event reports, particularly deaths submitted by health care professionals and consumers.
The increases seen among the younger women may be related to stimulated reporting by health care professionals for TT events that are uncommon occurrences among young women. The reports of cerebrovascular events and myocardial infarctions associated with ORTHO EVRA in young women are particularly concerning. They warrant review by a neurologist and cardiologist to further evaluated the clinical aspects of the cases.
SUMMARY/CONCLUSION
During the period covered by this consult, all combined oral contraceptive products continue to accrue thromboembolic and thrombotic adverse event reports in AERS, with PE being the most frequently reported event. Only Ortho Evra, Yasmin and NuvaRing have reports for all four of the groupings: PE, DVT, CVE and OTEs. However, only Ortho Evra had fatal reports in the U.S. of cerebrovascular events and myocardial infarction. Of particular concern, the Ortho Evra patients experiencing cerebrovascular events and myocardial infarction were younger and not significantly heavier (with the exception of one morbidly obese patient), when compared to the groups.
The differences in reporting rates are particularly noticeable for women younger than 30 years of age.
The reports of cerebrovascular events and myocardial infarctions associated with Ortho Evra in young women are particularly concerning…
The increased number of reports from stimulated reporting, however, provides an opportunity to better characterize the events experienced by individual women adversely impacted by the use of these products.
j. ORTHO EVRA PATCH NO ADVANTAGE OVER BIRTH CONTROL PILLS FOR PREVENTION OF TEEN PREGNANCY
(APPENDIX 6 ¶ 546-548)
204. A. Thurman, et al,( Preventing Repeat Teen Pregnancy: Postpartum Depot Medroxyprogesterone Acetate, Oral Contraceptive Pills, of the Patch ? J. Pediatr Adolescent Gynecol (2007) 20:61-65) prospectively evaluated repeat teen pregnancy rates within one year of delivery, among adolescent women who chose to use the contraceptive patch (ORTHO EVRA)(n=55) versus oral contraceptive pills ("OCP")(n=55) versus Depo Provera (n=142) for postpartum contraception. Participating postpartum teens were 11-19 years of age; 72% were African American and 96% qualified for Medicaid. A telephone interview was conducted every 3 months. The primary outcome was repeat pregnancy within 12 months of the index delivery. Secondary outcome variables were continuation rates of a hormonal contraception methods, reasons for discontinuation of a hormonal contraceptive method, side effects and condom usage. At 1-year follow-up, repeat pregnancy rates were 14.2%, 29.7% and 31.8% among Depo Provera, OCP and patch users respectively (P=0.02). Depo Provera users were more likely to be using any form of hormonal contraception I year postpartum than either patch or OCP users. Condom use was similarly low among all cohorts. Of adolescents who give birth, 17-35% will become pregnant again within 1 year of delivery.
205. From the tables, of the 55 women started on the ORTHO EVRA patch, 12 (22%) were age 13-16; 26 (47%) were ages 17-18 and 17 (31%) were 19 years of age. At 3 months, 35 (64%) women were still using the patch and not pregnant, 4 (7%) had changed to another hormonal method, and 10 (18%) were not using a hormonal method, accounting for 49 women (89%). At the conclusion of the study at one year, 31.8% (14 women) of the 44 women with follow-up started on the patch were pregnant. There is additional information for 29 women begun on the patch who were not pregnant at one year providing a follow-up at one-year for 44 women (80%) begun on the patch. At one year for 44 women, 14 women (32%) have remained on the patch and were not pregnant, 2 (5%) stopped the patch and began a different hormonal method, and 13 (30%) were not using a hormonal method, and 11 were lost to follow up (20%). For the OCP, there was one-year data for 37 women, with 11 (30%) women pregnant at one year. Of the 37 OCP women, 13 (35%) were remaining on OCPs at one year, 6 (16%) using another method and 6 (16%) using no hormonal method, and 18(33%) lost to follow-up. For the adolescent post delivery population in a prospective one years study, there was no significant difference between the outcomes for OCPs versus the ORTHO EVRA patch in terms of compliance and prevention of unintended pregnancy.
206. The authors wrote:
In the United States, adolescents most commonly choose oral contraceptives pills (OCP) for birth control. However, teens have high contraceptive failure rates, and difficulty with every day dosing. The contraceptive patch (ORTHO EVRA) was approved by the Federal Drug Administration (FDA) in 2001 and its convenient, once-weekly dosing regimen is heavily marketed. In adults, the contraceptive patch has improved dosing compliance when compared to OCPs but longitudinal data on its efficacy in postpartum teens is lacking.
Our data suggest that the patch offers no advantage over OCP in terms of preventing repeat teen pregnancies within I year. Our data support that long-term, effective contraception, such as DMPA, is likely part of the solution to the problem of recurrent teen pregnancy.
k. ORTHO EVRA in a HIGH RISK POPULATION (APPENDIX 6 ¶526-528)
207. A. Bakhru and N. Stanwood, (Performance of Contraceptive patch Compared with Oral Contraceptive Pill in a High-Risk Population, Obstetrics & Gynecology, Vol. 108, No.2, August 2006, pp 378-386.) performed a prospective assessment of the contraceptive patch and oral contraceptive pill in a population at high risk for future unintended pregnancies and abortions. A total of 1,230 contraceptive-nave women were identified from women undergoing contraceptive counseling at three Planned Parenthood Clinics. The primary outcome measure was time to discontinuation of the patch or pill. Subjects were racially diverse and primarily single women. Loss to follow-up was higher among patch users (45.2% versus 29.5%, p<.001). Verified continued use beyond the first three cycles was lower with patch users (67% versus 89%, P<.001). Skin irritation and site reactions were the treatment-limiting factor for 3.3% of patch users. The 3,206 cycles captured in the study resulted in a Pearl Index of 3.62 for the pill and 14.84 for the patch.
208. Of the patch users, 26% were 17 years or less, 35.8% were 18-22 years and 38.3% were 23 or greater, 86.2% were single, 52.7% were Caucasian, 35.4% were African American, 23.2% had one prior birth, 33.3% had one prior abortion, 19.2% had no formal birth control method, 38.7% used condoms alone, and 25.1% had been pregnant prior to starting contraception. In contrast, for the OCP users, 42.7% were 17 years or less, 30.9% were 18-22 years and 26.4% were 23 or older, with 88.1% single, 82.9% Caucasian, and 10.5% African American, 4.2% had one prior birth, 8.6% had one prior abortion, 19.5% had no prior birth control method, 60.4% used condoms, and 3.8% had been pregnant prior to starting contraception.
209. Among patch discontinuations, pregnancy, contraceptive preferences, skin reactions, compliance issues and increased or irregular vaginal bleeding were the most common reasons to discontinue the use of the ORTHO EVRA patch. According to the authors:
By life tables survival analysis, pill users had a significantly higher continuation rate than patch users (P=.028)…At 1 year of follow-up, 76% of pill users and 57% of patch users continued their method (P=.004). Among patch users, those with private insurance fared the best, followed by those self-paying, those in the Family Planning Benefit Program, and lastly those in Medicaid. Full-time workers also did better than part-time workers.
Among patch users, the multivariate Cox proportional hazards showed no difference in continuation by clinical location (P=.865), black versus white race (P=. 199), marital status (P=.652), high-risk status (P=.098), age (P=. 109), parity (P=.887), or abortion history (P=.209). Although those at high risk for a future unintended pregnancy demonstrated lower continuation rates than those not at high risk (P=. 16), in the multivariate model this was not significant and did not alter the findings. In the multivariate model, only work status was a significant predictor of discontinuance (P=.033)
DISCUSSION
In this study, we found the patch to have significantly higher discontinuation and unintended pregnancy rates in this high-risk population. We hypothesized that the patch would be equal or superior to the contraceptive pill in this population. The patch claims simplicity, reversibility, and the convenience of weekly dosing with a generous fail-safe period, providing a defense against imperfect dosing. We believed any concerns with the contraceptive patch would be minimal compared with expected increases in acceptability, tolerance, and compliance. However, we found the opposite.
Our continuation rate for the patch, 57% at 1 year, represents the first report of typical use continuation in the literature. The lower continuation rate for the patch may be due, in part, to a lower baseline familiarity with transdermal delivery system, given its introduction in 2001, compared with the pill introduced in 1960.
We were also interested in the tolerability of the patch, given previous reports of skin irritation for those using the patch. Skin irritation or site reactions were cited as the treatment-limiting factor by only 3.3% of patch users. It may have been a deterrent from perfect compliance in many more users, however, prior trials have reported skin site reactions in 4.7% of users and treatment-limiting skin reactions in only 2.6% of users. Additionally, although the patch was tested in humid climates, it has not been evaluated in colder and drier northern climates such as Western New York. In such climates, widespread use of lotions for dry skin may reduce the effectiveness of the patch. It is unclear if lotion use and drier climate was a factor in our poorer outcomes with the patch. It has not been investigated to our knowledge.
Other factors may explain the discrepancy between the literature and our results.
…
Our study disproved our initial hypothesis that the patch would have a better continuation rate than the pill in this high-risk population. We found the contraceptive patch to have lower rates of continuation and effectiveness, we could not, however, discern the difference between user failure and method failure. Our findings do not discount use of the contraceptive patch for high-risk population. Further research should investigate reasons for poorer real-world performance of the patch and investigate ways to assist women who are at high risk for unintended pregnancy in using the patch effectively.
OPINION #5:
210. By J&.J's own admission in internal documents, J&J placed corporate profits over patient safety and the need for adequate warnings in its label. J&J's disregard directly contributed to patient injuries and death.
ADDITIONAL SUPPORT FOR OPINION #5
a. IMPACT OF LABELING CHANGE (POE 00722692):
• Hypothetically, if PK results were to indicate EE absorption falls out of the 80-125% range allowed, and DOE fails to bring EE into label specs…
-Commercial product would he considered >35mcg (40-45 mcg)
-Documented category risk threshold <35 mcg.
-FDA may require, and physicians will:
•USE ONLY 2nd LINE WHEN LOWER DOSE CONTRACEPTIVES FAIL
-New ORTHO EVRA forecast if 2nd line:
•2004: BP=$358M; If 2nd line by 4Q04= $270M
•2005: SP=$478M; If 2nd line = $48M*(assume 10% of remaining SP 2006).
211. With the release of the increased EE delivery information by the FDA and in the label and greater awareness by the consumers of the increased risks from the media, the sales of ORTHO EVRA began to plummet. An example of decreased sales is a December 2005 email from Christi Shaw to Michelle Brennan: Comparison of Consumer Weekly Tracker to Consumer Analog. There is a table for ORTHO EVRA Forecasts: (POE04898153). ORTHO EVRA consumer "continuing business" of 80% were analyzed as: Socially unsettled (41%) [60% awareness]- 50% will quit, 30% will call doctor and, of those, 70% will quit; liberated socialite (29%) [80% awareness]- 20% will quit, 70% will call doctor and 80% will then quit; and reproductive relief seeker (11%)[80% awareness]- 20% will quit, 60% will call doctor and then 70% will quit. Ms. Shaw wrote the following regarding sales:
Can't sleep so I whipped up this comparison based on the consumer tracking we received today…
Based on the first two weeks of data:
I. Awareness (61%) after only two weeks is almost at the level we predicted would occur by three months. 2. Lost 32% of business based on second weekly compared to 50% we predicted after 3 months in consumer analog. 3. While I do think awareness will continue to increase and very well could hit 80%* by three months, I am not sure consumer action will change dramatically as the most noise was within the first two weeks.
So let's assume that awareness overall goes to 80% and consumer action stays relatively stable (let's assume a combined quit or switch) then it looks like the consumer analog would suggest a 44% loss in our continuing business after 3 months versus the 50% suggested. In fact, if you want to assume 90% awareness after 3 months then the loss would be 50% (POE 04898152).
b. Dr. CAUBEL's RESIGNATION LETTER TO J&J OCTOBER 2005(APPENDIX 6 ¶465-466)
212. There was an October 31, 2005 letter of resignation from Dr. Patrick Caubel, to Mr. William Weldon, COB J&J (Exhibit 179). He had left his position as Vice President, Therapeutic Area leader in Benefit and Risk Management after more than 7 years. Joseph Scodari, Worldwide Chairman of the Pharmaceutics Group, J&J, testified that he had met with an executive who expressed his frustration over the fact "that young women were dying of massive pulmonary embolisms" and "that there was not enough sense of urgency" (Scodari, 10/6/06. p 152-155, Exhibit 179). Dr. Caubel, the executive referenced by Mr. Scodari, sent a letter to Mr. Weldon (Exhibit 179) to clarify his reasons for his leaving J&J.
213. In Dr. Caubel's position, he had been charged with benefit/ risk management and safety evaluations, and he felt that he had not been able to exercise his responsibilities. As he discussed below, he had been involved with the safety evaluation of ORTHO EVRA since its launch on the market in the US. He had been the medical monitor for NED-I that led to the discovery of 60% increased delivery of EE, which was higher with the patch when compared to COC. Shortly after the product launch, "there was an unusually high number of thromboembolic accidents started to be reported to the Company". A preliminary epidemiologic study (the SS. Jick study) led to the conclusion that the risk of venous thrombosis was not increased with ORITHO-EVRA compared to OCs. He indicated he used the word "preliminary" because these findings were only partial results of one of two on-going studies, the other study being the Ingenix (i3) study. He was aware in October 2005 of more than twenty fatalities that had been reported for women using ORTHO EVRA. He was also aware that J&J decided to quickly try to communicate the partial and incomplete favorable epidemiology study results despite:
…the compelling evidence that the estrogenic exposure was unusually high compared to the oral product, and that the reporting rate of fatalities and over various thromboembolic disorders was out of range for this class of products.
OPINION #6
214. The J&J IND application for ORTHO EVRA clearly described to the FDA that J&J intended to market a transdermal contraceptive system patch with increased patient safety, improved patient compliance and reduced estrogen exposure. The FDA approved NDA 21-180 on November 2001 for J&J to market a combination contraceptive patch that was capable of delivering approximately 150mcg norelgestromin and approximately 20 mcg ethinyl estradiol and was comparable to the absorbed doses of hormones following oral contraceptive administration of 250 mcg norestimate and 35 mcg ethinyl estradiol (POE 0026998). First at a June 29, 2004 meeting, followed by a July 20, 2004 letter, and again at an August 18, 2005 meeting, the FDA requested that J&J voluntarily revise its ORTHO EVRA label to remove reference to 20 mcg EE delivery information since it was "misleading" for physicians and women (APPENDIX 6 ¶ 445, 349-351) (POE 001027767-72)). From the data reviewed, J&J deviated from the behavior of a responsible pharmaceutical manufacturer when it began to commercially market an ORTHO EVRA product despite awareness that its ORTHO EVRA patch did not perform in women as originally intended, designed and approved by FDA.
ADDITIONAL SUPPORT FOR OPINION #6
a. IND 50,488 TRANSDERMAL CONTRACEPTIVE SYSTEM (TCS) -APRIL 30, 1996
215. In J&J's IND, within a discussion of its proposed clinical protocols for the contraceptive patch product (POE0000026), J&J described its rationale for seeking to market a TCS patch. It provided the FDA with reasons why it should want to approve protocols for obtaining human clinical trial data for prevention of pregnancy based on public health considerations (*Underlining added for emphasis):
The 17d-NGM/EE TCS could be an innovative approach to contraception: It has the potential to deliver effective contraception utilizing a lower dose of steroids than oral contraceptives (OCs) and may be more convenient to use than OCs. The non-oral delivery of a contraceptive steroid is desirable because the hepatic elimination of much of the steroid via the first pass effect is avoided. Consequently, a reduced dose of steroid may be sufficient to achieve serum steroid levels comparable to those produced by current OCs. In addition, some of the OC side effects related to ingestion (nausea) and hepatic metabolism ( enzyme induction, adverse lipid and coagulation factor changes) are likely to be decreased. It is likely that an estrogen dose lower than OCs will be delivered by the patch with a comparable efficacy. This could be a clinically important advance in safety by potentially reducing the risk of thromboembolic events.
A seven-day TCS will theoretically have a higher rate of compliance than an OC because it will be administered regularly on a weekly, rather than a daily basis. A substantially greater rate of compliance with patches than with oral administration has been demonstrated for several different kinds of patches. For women who do not always take OCs correctly or who have side effects associated with peak serum levels of steroids, the contraceptive patch would be the most logical alternative contraceptive method before turning to long-acting progestin-only contraceptives.
In addition, a combination estrogen/progestin contraceptive is necessary to avoid irregular menstrual patterns, which is the primary reason for discontinuation of the progestin-only delivery systems. The 17d-NGMN/EE TCS is expected to produce cycle control profile equal or better than that of OCs and clearly superior to progestin-only pills, injectables and implants.
216. The J&J IND application clearly described to the FDA that J&J intended to test an investigational TCS patch that had a potential for increased safety, patient compliance and "reduced estrogen exposure". J&J proposed a clinical investigation plan based on the outcome of its completed Phase 1 study (NRGEEP-PHI-003) (all patches applied to the abdomen) that demonstrated feasibility of the design concept- the patch could be worn safely by 12 healthy women for one week and that steady state profiles of both EE and 17d-NGM could be established.
b. END OF PHASE 2 MEETING WITH THE FDA – JULY 9, 1997
(APPENDIX 6 ¶ 111-117)
217. According to a September 3, 1997 FDA letter from Heidi Jolson, MD, MPH, Deputy Director, Division of Reproductive and Urologic Drug Products, an End of Phase 2 Meeting was held between the FDA and J&J on July 9, 1997 (POE 00013856). The proposed commitments from J&J and the FDA included conducting three efficacy studies- a US comparative trial with Triphasil, a European comparator trial with Mercilon (a product not available in the US) and a Global comparator trial. The US and European trials would be pivotal and the Global trial supportive. The study protocols in all three trials were to be identical. 1600 women would be enrolled in 40 US sites, non-US trials would be performed under ICH conditions and would also be submitted to the NDA. The sponsor should conduct the post-study follow-up by scheduling the last visit no less than 14 days after the last active patch was used. The sponsor was strongly encouraged to keep all women enrolled and accounted for in the study data including those with less than optimal compliance.
218. The decisions reached included that: because Mercilon was not approved in the US, comparative claims based on the European study could not be made; areas of proposed claims of superiority must be defined in advance of the study; and because the US intent to treat population was not powered to make comparison with Triphasil, the protocol should state that only descriptive analyses will be used. The proposed clinical pharmacology and biopharmaceutics program was deemed acceptable; however, the sponsor was encouraged to address missed or detached patches via simulation or modeling.
219. At the end of Phase 2 meeting, Dr. Shangold presented the Phase 2 clinical study results to the FDA. Dr. Dorantes, FDA, inquired how J&J evaluated patch adhesion. Dr. Shangold replied that if less than 50% of the patch detached, it was smoothed back in place. If greater than 50% detached, a new patch was to be used. Dr. Dorantes advised that sponsors of other transdermal products scored detachment on a 25, 50, 75% scale. Dr. Shangold noted that approximately 30 of 12,000 TCS had become detached during Phase I and 2 studies and 2/3's occurred in the first cycle.
220. Dr. Abrams reviewed the pharmacokinetic (PK) results to date for the FDA and the planned human PK studies. Dr. Mauck, FDA, questioned if the TCS really delivered a "reduced steroid dose", since the amount of 17-dacetylnorgestiamte is the same as seen after oral administration. Dr. Abrams replied that TCS serum levels are comparable to those seen after oral administration but liver-first pass metabolism is avoided, gastrointestinal effects should he less, and the reduction of the peak/trough effect seen after oral administration should result in better metabolic and safety profile for the TCS. He added, however, "this needs to be proven". Additionally, to achieve these levels of ethinyl estradiol orally (??25ug/day) requires the administration of 35ug daily for a pill, most of which is rapidly presented to the liver. Dr. Mauck then questioned if TCS serum concentrations would be higher than OCs. Data was presented to show that this was not the case. Dr. Dorantes questioned the proportion of drug load released from the patch after 7 days. Ms. Karp, J&J, noted it was approximately 50%. Dr. Dorantes also provided insight that there were guidelines for changing a manufacturing site and indicated that a standard bioequivalence study may not be required.
221. Dr. Shangold presented the proposed Phase 3 pivotal studies. Dr. H. Jolson, FDA, asked how patients would be reminded to change the patch. She was advised that there were several approaches being contemplated. Dr. Rarrick, FDA, insisted that data from the last cycle be included in the efficacy evaluation and PRI agreed to consider moving the data of post-study pregnancy evaluation.
222. Dr. Mauck asked if J&J was considering different color patches. She was advised this was a long term goal. Mr. Corrado questioned requirements for a colored overlay. Dr. Mauck confirmed that a PK study would be required.
223. Dr. Kammerman, FDA, noted that in the Triphasil protocol there had been no power calculations for contraceptive efficacy endpoints. Dr. Rubin, J&J, responded that there is not enough power to show equivalence or superiority for the contraceptive efficacy endpoints. The primary endpoint is a point estimate, the Pearl Index. Dr. Kammerman continued with the need for an adjustment if J&J wants to make claims about patient compliance and cycle control, since they are multiple outcomes. Dr. Dorantes suggested that J&J should also determine intra-subject variances in the multiple dose PK study since we will be looking at serum concentration data during 3 consecutive weeks of wear during Cycle 3. J&J agreed.
224. I reserve the right to amend my opinions.