• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

For Consumers

  • Print
  • Share
  • E-mail

FDAMA Section 113: Status Report on Implementation (August 2005)

TABLE OF CONTENTS

EXECUTIVE SUMMARY

INTRODUCTION

CHAPTER 1. BACKGROUND
I. ClinicalTrials.gov - the Clinical Trials Data Bank
II. Guidance Documents

CHAPTER 2. EDUCATION PROGRAM 
I. Objectives
II. Methods
A. CDER
B. CBER 
III. Results
IV. Discussion

CHAPTER 3. COMPLIANCE EVALUATION PROGRAM I
I. Objective
II. Methods
A. Data collection 
B. Data extraction 
C. Quality control 
D. Documenting trial listings in ClinicalTrials.gov 
III. Results
A. New protocols submitted to CDER and meeting the guidance criteria for listing in ClinicalTrials.gov 
B. Required trial listings in ClinicalTrials.gov 
C. Voluntary trial listings in ClinicalTrials.gov 
D. Sponsor trial listings in ClinicalTrials.gov 
E. Trial listings in ClinicalTrials.gov sorted by FDA review divisions 
IV. Discussion
A. Definition of timing for submissions
B. Identification of serious diseases

CHAPTER 4. COMPLIANCE EVALUATION PROGRAM II 
I. Objectives
II. Methods
A. Data collection
B. Data extraction 
C. Documenting trial listings in ClinicalTrials.gov 
D. Follow-up Telephone Calls
III. Results
A. New protocols submitted to the Division of Oncology Drug Products and meeting the guidance criteria for listing in ClinicalTrials.gov.
B. Protocols by sponsor type 
C. Trial listings in ClinicalTrials.gov by sponsor type 
D. Comparison of 2002 versus 2004 cancer data 
E. Follow-up telephone calls
IV. Discussion
A. Business decisions to stop or delay funding for the development of a specific drug 
B. Lack of knowledge about the law and/or guidance 
C. Identification of trials to test effectiveness 
D. Concerns about availability of information to the public and to competitors

CHAPTER 5: PUBLIC AVAILABILITY OF INFORMATION
I. Trials Not Listed in ClinicalTrials.gov
A. Objective 
B. Methods 
C. Results
D. Discussion 
II. Trials Listed in ClinicalTrials.gov with Limited Information about Location and Contact Information, and Drug and Sponsor Name
A. Location and Contact Information
B. Drug Name and Sponsor Name 
C. 2004 Update

CHAPTER 6. REPORTING CLINICAL TRIAL RESULTS IN THE CLINICAL TRIALS DATA BANK 
I. Objectives
II. Methods
A. References
B. Links
III. Results
A. References
B. Links

CHAPTER 7. LIMITATIONS

CHAPTER 8. SUMMARY

APPENDICES

Appendix A: Acronyms & Definitions

Appendix B: CDER Letter Template

Appendix C: CDER Education Program

Appendix D: CBER Letter Template

Appendix E: CBER Education Program

Appendix F: Compliance Evaluation Program I: Data Collection Process

Appendix G: Protocol Form 2

Appendix H: List of Serious and Non-Serious Diseases and Conditions

Appendix I: Compliance Evaluation Program I: Additional Data Elements

Appendix J: Data Elements Extracted for Office of Women's Health Project

Appendix K: Office of Women's Health Data Entry Form

Appendix L: Letter Results Data Entry Form

Appendix M: Thank You Letter

Appendix N: Compliance Evaluation Program II: Data Collection Process

Appendix O: Submissions Form

Appendix P: Trial Site Location Letter

Appendix Q: Document Check-In Data Entry Form

Printer-friendly PDF version of this Report (683 KB)

 INTRODUCTION

Time is a precious commodity for many patients who run out of standard treatment options and want to explore participation in a clinical trial. Learning what trials are being conducted, where they are taking place, and who is eligible to participate can be a challenge, especially if a patient has a serious or life-threatening disease.

At the same time, recruiting patients to clinical trials is time consuming and costly for the pharmaceutical industry. Everyone -- clinical trial sponsors, patients, and patient advocates -- has an interest in facilitating the studies to evaluate new treatments for serious illnesses.

In 1997, the U.S. Congress addressed the needs of patients and their advocates by adding a provision to the Food and Drug Administration Modernization Act (FDAMA) that mandated the establishment by the National Institutes of Health (NIH) of a publicly-accessible clinical trials data bank and required sponsors to list eligible trials in the data bank. The NIH, through its National Library of Medicine (NLM), developed the Clinical Trials Data Bank called ClinicalTrials.gov. This service was made available to the public in February 2000 at www.clinicaltrials.gov.

In January 2002, the Food and Drug Administration (FDA) Office of Special Health Issues (OSHI) undertook a study to investigate listings of study protocols in the Clinical Trials Data Bank. The effort resulted in multiple projects on various issues related to the topic.

The first two projects were intended to educate private-sector sponsors about the statutory reporting requirements under Section 113 of FDAMA and to assess sponsor compliance with the law. The results of these projects stimulated questions about whether the number of clinical trial listings submitted by sponsors to the clinical trials databank was increasing over time. Consequently, in May 2004 another project was initiated to further assess compliance and, if possible, identify sponsors' reasons for not complying. Recent attention to publicly available clinical trial results prompted another project that reviewed what additional information sponsors were voluntarily including in their submissions to the data bank. This report will examine each of these issues.

HOW THIS REPORT IS ORGANIZED

This report describes our key findings and is structured as follows:

  • Chapter 1: Background explains the history of the Clinical Trials Data Bank and the FDA guidance documents implementing the requirements of the data bank.
  • Chapter 2: Education Program describes the objectives, methods, and results of the project intended to educate private-sector investigational new drug (IND) sponsors about the statutory requirements, FDA guidance documents, and web-based entry tool for the Clinical Trials Data Bank.
  • Chapter 3: Compliance Evaluation Program I describes objectives, methods, and results for the part of the project in which we examined the extent to which trials submitted to the Center for Drug Evaluation and Research (CDER) that were required to be submitted to the Clinical Trials Data Bank were actually submitted to the data bank.
  • Chapter 4: Compliance Evaluation Program II describes the objectives, methods, and results for the project that investigated whether there has been an increase in the number of cancer trials submitted to the Clinical Trials Data Bank. In this section, we also discuss some additional reasons why we believe sponsors may not have listed trials on the data bank.
  • Chapter 5: Public Availability of Information explores the degree to which information not submitted to the Clinical Trials Data Bank was found on other publicly available resources.
  • Chapter 6: Reporting Clinical Trials Results discusses the objectives, methods, and results for the project that investigated what additional information with respect to clinical trial results sponsors were voluntarily submitting to the Clinical Trials Data Bank.
  • Chapter 7: Limitations discusses the limitations of the various projects.
  • Chapter 8: Summary provides the current status of implementation for FDAMA Section 113 and suggests some considerations for the future.
  • Glossary: See Appendix A for a list of key acronyms and definitions contained in the report.

 CHAPTER 1. BACKGROUND

I. ClinicalTrials.gov - the Clinical Trials Data Bank

Section 113 of FDAMA (42 U.S.C. § 282(j)), enacted November 21, 1997, creates a public resource for information on studies of drugs, including biological drug products, to treat serious or life-threatening diseases and conditions, conducted under the FDA's IND regulations at Title 21, Code of Federal Regulations Part 312 (21 CFR 312). It directs the Secretary of Health and Human Services (HHS), acting through the Director of the NIH, to establish, maintain, and operate a data bank of information on clinical trials.

Specifically, Section 113 of FDAMA requires that the Clinical Trials Data Bank contain:

  1. information about both federally- and privately-funded clinical trials for experimental treatments (drug and biological products) for patients with serious or life-threatening diseases and conditions;
  2. a description of the purpose of each experimental drug;
  3. patient eligibility criteria;
  4. a description of the location of clinical trial sites; and
  5. a point of contact for patients wanting to enroll in the trial.

Section 113 of FDAMA requires that information be forwarded to the data bank by the sponsor of the trial not later than 21 days after the approval of the protocol.1 Information provided through the Clinical Trials Data Bank is required to be in a form that can be readily understood by the public [42 U.S.C. 282(j)(3)(A)].

The NIH, through its National Library of Medicine, and with input from the FDA and others, developed the Clinical Trials Data Bank. The first version of the Clinical Trials Data Bank was made available to the public on February 29, 2000, via the Internet at www.clinicaltrials.gov. At first, the data bank, referred to as ClinicalTrials.gov, contained more than 4,000 records representing primarily trials sponsored by the NIH. Today, ClinicalTrials.gov includes more than 13,500 federally- and privately-sponsored trials.

II. Guidance Documents

FDA issued two draft guidances, a final guidance, and another draft guidance to address statutory and procedural issues related to Section 113 of FDAMA. Details of these guidance documents are described below.

March 29, 2000 Federal Register (65 FR 16620)
Draft Guidance for Industry: Information Program on Clinical Trials for Serious or Life-Threatening Diseases: Establishment of a Data Bank.2

  • Provided the pharmaceutical industry and other IND sponsors with recommendations on submitting trial information to the Clinical Trials Data Bank.
  • Included information about the types of clinical trials for which submissions to the Clinical Trials Data Bank are required under FDAMA Section 113 as well as the content of those submissions.

July 9, 2001 Federal Register (66 FR 35798)
Draft Guidance for Industry: Information Program on Clinical Trials for Serious or Life-Threatening Diseases: Implementation Plan.3

  • Addressed procedural issues, including how to submit required and voluntary trial information to the Clinical Trials Data Bank, and requests for an exemption to list a trial.
  • Proposed a time frame for submitting the information.

March 18, 2002 Federal Register (67 FR 12022)
Final Guidance for Industry: Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions.4

  • Combined the two draft guidances into a single document and incorporated comments received on the draft guidances.

January 27, 2004 Federal Register (69 FR 3923)
Draft Guidance for Industry: Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions.5

  • Revised the March 2002 guidance to include guidance for sponsors who will be submitting information required by the Best Pharmaceuticals for Children Act (BPCA).
  • The new information required by the BPCA includes a description of whether, and through what procedure, the sponsor of the research will respond to requests for access to the therapy outside of the clinical trial setting, particularly in children.

 CHAPTER 2. EDUCATION PROGRAM

This chapter describes the project intended to educate private-sector IND sponsors about the Clinical Trials Data Bank.

I.  Objectives

The objectives of the Education Program were to inform private-sector IND sponsors about the following:

  1. Statutory requirements for the Clinical Trials Data Bank under FDAMA Section 113
  2. The 2002 guidance Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions
  3. The availability of the Protocol Registration System (PRS), a web-based data entry tool
II. Methods

The education program consisted of mailing letters to IND sponsors of products regulated in CDER and the Center for Biologics Evaluation and Research (CBER) that served to inform each sponsor about FDAMA Section 113 requirements, the availability of the final guidance to industry, and the PRS.

Due to resource limitations, we only sent letters to "commercial" IND sponsors.6 We did not send letters to NIH, even for protocols classified as commercial, because the NIH institutes were already listing trials in ClinicalTrials.gov using procedures that had been developed prior to issuance of FDA's guidances.

A. CDER

A letter was mailed to the sponsor each time it submitted a new commercial protocol to a CDER IND (see appendix B for a copy of the CDER letter). Letters were created on a weekly basis throughout the duration of Compliance Evaluation Program I, from February through November 2002.

For a more detailed description of the process used to prepare the CDER letters, see appendix C.

B. CBER

Due to resource limitations, we did not obtain individual CBER protocols on an on-going basis. Rather, we obtained a list identifying each active CBER IND. Active INDs were grouped by sponsor and company contact person. Each company contact person was sent the same letter that CDER sponsors received but instead of referencing a specific protocol, it contained a list of that sponsor's active INDs (see appendix D for a copy of the CBER letter).

For a more detailed description of the process used to prepare the CBER letters, see appendix E.

III. Results

For CDER INDs, 1748 letters were mailed to sponsors of 1012 commercial INDs over the ten-month period of February to November 2002.

For CBER INDs, 612 letters were mailed to sponsors of 1388 commercial INDs during May 2002.

IV. Discussion

We did not request a formal response from each sponsor. On occasion we received written correspondence from a sponsor acknowledging the letter and stating the sponsor would review the referenced protocol to determine whether it met the listing criteria. If the protocol met the listing criteria, the sponsor would list the trial in ClinicalTrials.gov.

Throughout the duration of the education program, we evaluated what proportion of protocols submitted to FDA that were required to be submitted to ClinicalTrials.gov had actually been submitted for inclusion. This evaluation will be further explored in the next chapter.

 CHAPTER 3. COMPLIANCE EVALUATION PROGRAM I

This project attempted to determine the extent to which sponsors' IND submissions to CDER were included in the Clinical Trials Data Bank. It compared the number of trials submitted to CDER and to ClinicalTrials.gov for a period of nine months. CBER protocols were not evaluated for this project.

I. Objective

The objective of Compliance Evaluation Program I was to assess regulatory compliance with Section 113 of FDAMA by comparing the number of trials listed in ClinicalTrials.gov to the number of commercial protocols submitted to CDER between January 1, 2002 and September 30, 2002.

II. Methods

Data for the first evaluation program were obtained from two primary sources within the agency: Center ORACLE Management Information System (COMIS), and paper protocol records submitted by IND sponsors. New commercial protocols submitted to CDER between January 1 and September 30, 2002 were identified from a search of the COMIS database. The COMIS search included both new protocols and new INDs because it was assumed each new IND submitted contained at least one protocol.

A. Data collection

Each week we received an electronic file containing each new commercial protocol submitted to CDER the previous week. The file was downloaded into a Microsoft Access™ database we developed for the project, hereafter referred to as the OSHI database. Each week the CDER document room staff delivered copies of the paper INDs. By project's end, a total of 1,865 paper INDs were delivered to OSHI.

For a complete description of the data collection process see appendix F.

B. Data extraction

The following key data elements were extracted from 2,062 protocols and entered into the OSHI database using the Protocol Form 2 (see appendix G for a copy of Form 2):

  • Indication: The indication provided by COMIS was revised using NLM's controlled vocabulary thesaurus known as Medical Subject Headings (MeSH) (http://www.nlm.nih.gov/mesh/MBrowser.html). MeSH was used to provide consistency in the indication data field and enabled us to accurately and efficiently calculate the number of protocols per disease.
  • Sponsor type: Industry (pharmaceutical company), NIH, Other Government, Physician, Medical Center, Other
  • Protocol number: Sponsor designated protocol identifier
  • Protocol title: Title of protocol
  • Number of protocols within the submission
  • Phase: 1, 1/2, 2, 2/3, 3, 3/4, 4, Not Specified (NS)

If the phase was not mentioned in the title or protocol synopsis then the phase listed on FDA Form 1571 was used. We used FDA Form 1571, a cover sheet for an IND submission, to identify the phase in approximately 10% of all protocols.

Although 21 CFR 312.21 states that a clinical investigation of a previously untested drug is divided into three phases (1, 2, 3), we included transitional phases (1/2, 2/3, 3/4) to reflect how sponsors categorize protocol submissions.

  • Is the protocol indicated for a serious or life-threatening disease or condition? Yes or No

Currently, FDA does not maintain a list of serious or life-threatening diseases and conditions. FDA has defined serious or life-threatening diseases and conditions in previous documents. Most recently, FDA discussed issues related to products intended to treat serious or life-threatening diseases and conditions in the guidance for industry on Fast Track Drug Development Programs -- Designation, Development, and Application Review (July 2004).7 In conjunction with the CDER Office of Medical Policy and directors of the CDER review divisions, we developed a list of serious and non-serious diseases and conditions for this project. The list includes 120 serious and 72 non-serious diseases and conditions (see appendix H for the list of serious and non-serious diseases and conditions). It was evident from discussions during this project that the seriousness of a disease is often a matter of judgment and can vary by protocol. For example, the seriousness of angina can vary greatly depending on whether it is stable or unstable, new-onset or chronic. For purposes of this project we made a determination for the entire project to ensure consistency (See Chapter 7 - Limitations). This list does not have official status for any purposes other than this project.

  • Does the study test effectiveness? Yes, No or NS. If the study listed efficacy endpoints as primary or secondary objectives then "yes."

For a complete list of data elements extracted for Compliance Evaluation Program I see appendix I.

We also extracted data for a project conducted by the Office of Women's Health (OWH) (see appendix J for a list of the data fields extracted for OWH). The data elements were entered into the database using the OWH Form (see appendix K for a copy of the OWH Form).

C. Quality control

Quality control was conducted on a random sample of 515 protocols, or one out of every four protocols entered into the database, to assure that no more than 10% of the records included an error in any field.

D. Documenting trial listings in ClinicalTrials.gov

NLM prepared weekly reports to provide an update on sponsor trial listings in ClinicalTrials.gov. These reports were used to verify whether IND trials submitted to CDER were being listed in ClinicalTrials.gov. They contained the following: the date the trial was released into ClinicalTrials.gov, sponsor name, official representative's name, IND number, serial number, sponsor protocol ID, NLM identifier, title, condition, and drug name. We used the NLM reports as follows:

  • The IND number, serial number, and sponsor protocol ID from the NLM report were compared to data in the OSHI database to verify whether the protocol had been submitted to CDER within the 9-month period of January 1 - September 30, 2002. The comparison process began on January 1, 2002 and continued until March 2004 in order to capture trials with delayed enrollment.
  • A final comparison was conducted in March 2004 to verify whether trials meeting the criteria for inclusion in ClinicalTrials.gov (Phase 2, 3, or 4 effectiveness trials for serious or life-threatening diseases and conditions) and contained in the OSHI database were actually listed in ClinicalTrials.gov.
  • If the ClinicalTrials.gov trial was in the OSHI database, then the date the trial was listed in ClinicalTrials.gov and the assigned NLM identifier were recorded in the OSHI database (see appendix L for a copy of the Letter Results Form).
  • A thank-you letter was e-mailed to the sponsor's official representative acknowledging the listing in ClinicalTrials.gov (see appendix M for a copy of the thank-you letter).
III. Results

Of the 2,062 protocols submitted to CDER from January 1, 2002 to September 30, 2002 and entered in the OSHI database, we identified 688 protocols that were trials to test effectiveness for serious and life-threatening diseases or conditions and thus met the criteria for inclusion in ClinicalTrials.gov. Sponsors listed 239 (35%) of these 688 trials in ClinicalTrials.gov. Sponsors also listed an additional 35 trials that did not meet the criteria for inclusion.

A. New protocols submitted to CDER and meeting the guidance criteria for listing in ClinicalTrials.gov

A total of 1,865 IND submissions containing 2,062 new commercial protocols were submitted to CDER between January 1, 2002 and September 30, 2002. Of the 2062 protocols submitted, 66% (1361) were for serious and life-threatening diseases or conditions. Of the 1361 protocols for serious and life-threatening diseases or conditions, 80% (1088) were phase 2, 3, or 4 trials. Of the 1088 protocols, 63% (688) had primary or secondary efficacy endpoints and thus met the guidance criteria for submission to ClinicalTrials.gov.

Figure 1 below describes the process used to determine how many protocols met the guidance criteria for listing in ClinicalTrials.gov.

 Figure 1. Number of Trials Listed in ClinicalTrials.gov

Figure 1. Number of Trials Listed in ClinicalTrials.gov

B. Required trial listings in ClinicalTrials.gov

Thirty-five percent (239/688) of trials that should have been listed in ClinicalTrials.gov, based on our interpretation of FDAMA Section 113 as discussed above, were actually listed. Notably, almost three-fourths of the required phase 3 trials were not listed. Table 1 illustrates within each phase, what proportion of protocols that should have been listed were actually listed in ClinicalTrials.gov.

 Table 1. Trial Listings in ClinicalTrials.gov by Phase
Phase Number of protocols submitted to CDER Number of protocols required to be listed in ClinicalTrials.gov Number of protocols listed in ClinicalTrials.gov Percent of required protocols listed by phase
2 494 314 139 44%
2/3 30 17 8 47%
3 555 303 78 26%
3/4 5 3 1 33%
4 119 51 13 25%
Total 1203 688 239 35%

C. Voluntary trial listings in ClinicalTrials.gov

Sponsors are encouraged to list phase 1 trials and/or trials for drugs that are not intended to treat serious or life-threatening diseases and/or are not intended to test effectiveness. Table 2 provides a summary of these voluntary listings by phase.

 Table 2. Voluntary Trial Listings in ClinicalTrials.gov by Phase
Phase Number of voluntary trial listings in ClinicalTrials.gov
1 10
1/2 4
2 9
2/3 0
3 8
3/4 1
4 1
Not Specified 2
Total 35

D. Sponsor trial listings in ClinicalTrials.gov

We counted the number of sponsors who submitted trials to ClinicalTrials.gov as follows: if the sponsor listed multiple trials at different times during the 9-month period, that sponsor was counted once. Using this approach, 381 sponsors submitted 2062 protocols to CDER during the nine-month period of January 1 - September 30, 2002. Of the 381 sponsors, 370 (97%) were pharmaceutical company sponsors, 3 (1%) were government sponsors (CDC, NIH, or U.S. Army), and 8 (2%) were other (physician or medical center) as shown in Figure 2 below.

 Figure 2. Protocols Submitted to CDER by Sponsor Type


Figure 2. Protocols Submitted to CDER by Sponsor Type

As noted in Table 3, about half of the sponsors (187 / 381) submitted at least one protocol that met the criteria for inclusion in ClinicalTrials.gov.

 Table 3. Number of New Protocol Submissions by Sponsor
Number of protocols submitted to CDER in 9-month period Number of sponsors Number of sponsors who submitted to CDER protocols meeting the criteria for listing in ClinicalTrials.gov
1-5 309 126
6-15 48 37
16-60 19 19
>60 5 5
Total 381 187

Of the protocols submitted to CDER that met the criteria for inclusion in ClinicalTrials.gov, NIH listed 90% of their protocols and industry sponsors listed 30% of their protocols. A summary of the findings is presented in Table 4.

 Table 4. Compliance by Sponsor Type
Sponsor Type Total number of protocols submitted to FDA Trials meeting inclusion criteria Protocols meeting inclusion criteria and listed in ClinicalTrials.gov
Industry (Pharmaceutical Company) 1934 (94%) 626 (32%) 185 (30%)
Medical Center / University 3 (<1%) 1 (33%) 0 (0%)
NIH 112 (5%) 58 (52%) 52 (90%)
Other Government 5 (<1%) 1 (20%) 0 (0%)
Physician 1 (<1%) 0 (0%) -----
Other 7 (<1%) 2 (29%) 2 (100%)
Total 2,062 688 239

E. Trial listings in ClinicalTrials.gov sorted by FDA review divisions

The number of trial listings varied by FDA review division. Table 5 illustrates the compliance rate for industry-sponsored trials within each FDA review division. For example, of the submissions to the Division of Antiviral Drug Products by pharmaceutical companies, 50% of the trials that met the criteria for inclusion were listed in ClinicalTrials.gov. For submissions to the Division of Reproductive and Urologic Drug Products, only 4% of the trials that met the criteria for inclusion were listed in ClinicalTrials.gov.

 Table 5. Compliance Rate within FDA Review Divisions for Industry-Sponsored Trials
FDA review division Trials meeting inclusion criteria Protocols meeting inclusion criteria and listed in ClinicalTrials.gov
Antiviral 32 16 (50%)
Oncology 127 61 (48%)
Anti-Infective 9 4 (44%)
Neuropharmacology 88 32 (36%)
Metabolic/Endocrine 55 15 (27%)
Cardio-Renal 59 14 (24%)
Special Pathogen/Immunologic 22 5 (23%)
Analgesic/Anti-inflammatory/Ophthalmologic 31 6 (19%)
Anesthetic/Critical Care/Addiction 39 6 (15%)
Gastrointestinal/Coagulation 32 4 (13%)
Dermatologic/Dental 21 2 (10%)
Pulmonary 30 2 (7%)
Reproductive/Urologic 28 1 (4%)
IV. Discussion

Pharmaceutical companies listed fewer trials than expected despite a federal law, the issuance of several FDA guidance documents, educational mailings to sponsors, and an easy-to-use web-based data entry tool. Thirty five percent (239/688) of protocols that met the guidance criteria for inclusion in ClinicalTrials.gov were actually listed.

We considered possible reasons why pharmaceutical companies may not have listed trials or may have listed only limited information in ClinicalTrials.gov. The following reasons might explain the low compliance rate:

A) Definition of timing for submissions.
B) Identification of "serious" diseases.
C) Lack of knowledge about the law and/or guidance.
D) Concerns about availability of information to the public and to competitors.
E) Business decisions to stop or delay funding for the development of a specific drug.
F) Identification of trials to test effectiveness.

The first two reasons will be discussed below. The remaining reasons will be discussed as they relate to Compliance Evaluation Program II in the next chapter.

A. Definition of timing for submissions

The March 2002 guidance Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions specified that information for existing ongoing trials be submitted to ClinicalTrials.gov within 45 days after the guidance published. Accordingly, ongoing trials meeting the inclusion criteria were to be submitted on or before the implementation date of May 2, 2002. This project included protocols submitted to FDA prior to the implementation date based on the assumption that many studies would still be ongoing on the date of implementation.

We compared pre-implementation data to post-implementation data to assess any differences between compliance rates. For cancer protocols submitted from January 1-May 1, 2002 (pre-implementation), 46% of the protocols submitted by industry to FDA that met the criteria for inclusion in ClinicalTrials.gov were listed in the data bank. Forty-nine percent of the cancer protocols submitted from May 2-September 30, 2002 (post-implementation) by industry that met the criteria for inclusion in ClinicalTrials.gov were listed in the data bank. Although these data are cancer-specific, the data suggest there was little difference in compliance before or after the implementation date of May 2, 2002.

B. Identification of "serious" diseases

FDA does not maintain a list of serious diseases and conditions to assist sponsors in deciding if a protocol should be listed in ClinicalTrials.gov. However, FDA has defined serious or life-threatening diseases and conditions in previous documents. For example, FDA recently discussed issues related to products intended to treat serious or life-threatening diseases and conditions in the guidance for industry on Fast Track Drug Development Programs -- Designation, Development, and Application Review (July 2004).8 During the study period, it is theoretically possible some companies mistakenly thought only protocols for AIDS and cancer drugs or fast track drugs were required to be listed in the data bank. Indeed, in the guidance Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions, fast track drugs are used as an example of the type of drugs that would be considered as intended to treat serious and life-threatening conditions. However, the FDA guidance on Section 113 does not state that only fast track drugs need to be included in the ClinicalTrials.gov data bank, and includes a discussion of other diseases or conditions that can be considered serious or life-threatening.9

During the study period, it was evident that at least some companies are listing trials for conditions other than AIDS and cancer. We reviewed non-federal trials submitted to ClinicalTrials.gov between May 21, 2001 and November 21, 2003 (n=1100). Figure 3 illustrates the variety of diseases, excluding HIV and cancer, for which non-federal sponsors have submitted trials to ClinicalTrials.gov. For example, 26 different non-federal sponsors listed at least one protocol for diabetes. Trials for depression were listed by 21 different sponsors.

 Figure 3. Diseases for which Non-Federal Sponsors have Submitted Protocols

Figure 3. Diseases for which Non-Federal Sponsors have Submitted Protocols

To determine other reasons why sponsors were not listing eligible protocols and to assess more recent compliance with Section 113 of FDAMA, we initiated another project in 2004. Our findings from that project will be presented in Chapter 4.

 CHAPTER 4. COMPLIANCE EVALUATION PROGRAM II

In this section, we investigate whether there has been an increase in listing cancer trials in the Clinical Trials Data Bank since 2002. We also investigate additional reasons why sponsors were not submitting trials to the data bank.

I. Objectives

The objectives of the Compliance Evaluation Program II are listed below:

  1. To determine the regulatory compliance by comparing the number of cancer trials listed in ClinicalTrials.gov to the number of cancer protocols submitted to CDER's Division of Oncology Drug Products (DODP) during the three-month period of May 1, 2004 and July 31, 2004.10
  2. To compare the number of cancer trials listed in ClinicalTrials.gov in 2002 versus
    2004.
  3. To use cancer trials as an example to help determine reasons in general why sponsors are not listing their trials in ClinicalTrials.gov.
II. Methods

New commercial protocols submitted to CDER's DODP between May 1 and July 31, 2004 were identified from a search of the COMIS database. We selected cancer protocols and the May-July timeframe for evaluation for three reasons: 1) cancer is undeniably a serious condition, 2) to update compliance data for cancer that was presented publicly at the April 2003 FDA Science Forum,11 and 3) two years had passed since the May 2, 2002 implementation date.

A. Data collection

Our data collection process in Compliance Evaluation Program II was similar to the process used in Compliance Evaluation Program I.

For a complete description of the data collection process see appendix N.

B. Data extraction

We used the same data fields in Compliance Evaluation Program II as in Compliance Evaluation Program I, plus the following data fields were added:

  • Did the study list primary efficacy endpoints? Yes, No or NS
  • Did the study list secondary efficacy endpoints? Yes, No or NS

If the study listed efficacy endpoints as either primary or secondary, the study was considered to test effectiveness.

  • Comments: Insert the name of the sponsor's regulatory affairs person (contact person) responsible for the protocol in the event of follow-up phone calls to the sponsor.

Data were entered into the OSHI database using the Submissions Form (see appendix O for a copy of the Submissions Form). Data extracted during this review were compared to data from May 1 to July 31, 2002.

C. Documenting trial listings in ClinicalTrials.gov

NLM prepared weekly reports to provide an update on sponsor trial listings in ClinicalTrials.gov. The reports were used to verify whether IND protocols submitted to CDER were listed in ClinicalTrials.gov. The NLM reports contained the date the trial was released into ClinicalTrials.gov, sponsor name, official representative, IND number, serial number, sponsor protocol ID, NLM identifier, title, condition, and drug name.

  • The IND number, serial number, and sponsor protocol ID from the NLM report were compared to data in the OSHI database to verify whether the protocol had been submitted to CDER within the 3-month period of May 1 - July 31, 2004. The comparison process began on May 1, 2004 and continued until December 31, 2004.
  • A final comparison was conducted in January 2005 to verify whether trials contained in the OSHI database met the criteria for inclusion and were listed in ClinicalTrials.gov.
  • If the protocol was listed in the OSHI database, the date the trial was listed in ClinicalTrials.gov and the assigned NLM identifier was recorded in the Submissions Form.

D. Follow-up Telephone Calls

Twelve follow-up telephone phone calls were made to sponsors who appeared not to have listed cancer trials that met the criteria for inclusion. The follow-up call was an attempt to gather information about why the sponsor had not listed the trial.

III. Results

A. New protocols submitted to the Division of Oncology Drug Products and meeting the guidance criteria for listing in ClinicalTrials.gov.

One hundred forty commercial protocols were submitted to CDER's DODP between May 1, 2004 and July 31, 2004. Of the 140 cancer protocols, 39% (55) met the criteria for inclusion in ClinicalTrials.gov. Of these 55 trials, sponsors listed 76% of them in ClinicalTrials.gov. Also, sponsors voluntarily listed 19 out of 85 (22%) trials that did not meet the criteria for inclusion.

Figure 4 below describes the process used to determine how many protocols met the guidance criteria for listing in ClinicalTrials.gov.

 Figure 4. Number of Trials Listed in ClinicalTrials.gov

Number of Trials listed in ClinicalTrials.gov

 

B. Protocols by sponsor type

Of the 140 cancer protocols submitted to DODP, 100 (71%) were sponsored by pharmaceutical companies, 38 (27%) by NIH, and 2 (1%) by Other*. These results are shown in Figure 5 below.

 Figure 5. Protocols Submitted to CDER by Sponsor Type

Figure 5. Protocols Submitted to CDER by Sponsor Type
* Percentage does not add to 100.0 because of rounding.


C. Trial listings in ClinicalTrials.gov by sponsor type

A summary of compliance by sponsor type is presented in Table 6. Of the 55 trials meeting the inclusion criteria, 42 (76%) were listed in ClinicalTrials.gov.

 Table 6. Compliance by Sponsor Type
Sponsor Type Total number of protocols submitted to DODP Trials meeting inclusion criteria Trials meeting inclusion criteria and listed in ClinicalTrials.gov
Industry (Pharmaceutical Company) 100 32 21 (66%)
NIH 38 21 20 (95%)
Other 2 2 1 (50%)
Total 140 55 42 (76%)

D. Comparison of 2002 versus 2004 cancer data

When cancer trials submitted in 2004 were compared to those submitted in 2002, compliance increased from 61% to 76%. Table 7 compares the compliance rates by sponsor type.

 Table 7. Cancer Trials Submitted to CDER's DODP and Listed in ClinicalTrials.gov (2002 vs 2004)
Sponsor # Cancer trials listed in ClinicalTrials.gov
Jan - Sept 2002
# Cancer trials listed in ClinicalTrials.gov
May - July 2002
# Cancer trials listed in ClinicalTrials.gov
May - July 2004
Pharmaceutical Company 61 / 127
(48%)
22 / 44
(50%)
21 / 32
(66%) 70%*
NIH/NCI 52 / 57
(91%)
17 / 20
(85%)
20 / 21
(95%)
Other 2 / 3
(66%)
1 / 2
(50%)
1 / 2
(50%) 100%**
Total 115 / 187
(61%)
40 / 66
(61%)
42/ 55
(76%)

* 70% includes trials that were terminated or had delayed enrollment as identified in follow-up telephone calls.

** 100% includes a National Cancer Institute (NCI) cooperative group trial that was submitted to NCI's Physician Data Query (PDQ). There can be a time delay of 4-5 weeks to allow for the trial information to be entered into PDQ then sent to ClinicalTrials.gov.

E. Follow-up telephone calls

Twelve follow-up telephone calls were made to eleven pharmaceutical companies (one company had two protocols) and one sponsor coded as "other" who had trials that met the criteria for inclusion yet could not be located in ClinicalTrials.gov. The following reasons for not listing trials were cited by sponsors:

  • Trial was delayed in getting posted due to a vacancy in the clinical study team leader position. (n=1)
  • Drug development was terminated. (n=1)
  • Enrollment was postponed until first quarter of 2005. (n=1)
  • Sponsor had concerns about releasing proprietary information to the public and to competitors. Releasing the information would be detrimental to a small biotech company. (n=1)
  • Listing of the trial was overlooked. Trial would be listed as a result of the phone call. (n=3)
  • Sponsor acquired the drug from another company. Current sponsor did not verify the listing of the trial by the original sponsor. (n=1)
  • Lack of knowledge about the law and guidance. (n=2)
  • Sponsor initially intended to have domestic and international trial sites. Domestic sites were never opened and international sites were never listed because accrual was completed by the time of the phone call. (n=1)

The two remaining trials appeared not to be listed in ClinicalTrials.gov for the following reasons:

  • A pharmaceutical company-sponsored trial that appeared not to be listed in ClinicalTrials.gov was in fact listed. The FDA investigator misinterpreted the data presented in the ClinicalTrials.gov record. This error occurred because the sponsor did not use the same protocol number in ClinicalTrials.gov as it did in its submission to FDA.
  • In another instance, an NCI cooperative group trial was not listed in ClinicalTrials.gov. However, upon follow-up the trial had been submitted through NCI's cancer trial listing, the Physician's Data Query (PDQ) found on Cancer.gov. There can be a time delay of four to five weeks to allow for the trial information to be entered into Cancer.gov then sent to ClinicalTrials.gov.
IV. Discussion

The telephone calls made during Compliance Evaluation Program II confirmed many of our initial thoughts as to why companies were not listing trials. We identified the following reasons, and they are discussed below:

  1. Business decisions to stop or delay funding for the development of a specific drug.
  2. Lack of knowledge about the law and/or guidance.
  3. Identification of trials to test effectiveness.
  4. Concerns about availability of information to the public and to competitors.

A. Business decisions to stop or delay funding for the development of a specific drug

We know from discussions with pharmaceutical company representatives that sponsor business decisions may result in a delayed initial trial enrollment date for a protocol submitted to FDA. Sometimes trials are never initiated because funding for the project was discontinued. We expect that these factors accounted for very few of the protocols not listed in Compliance Evaluation Program I because we continued to review trial listings in ClinicalTrials.gov through March 2004, 18 months after we stopped receiving IND submissions. Any delayed trial listings were recorded in the OSHI database and were included in the overall compliance rate. Additionally, we know from the review of 2004 oncology data that only 9% (1/11) of trials not listed could be accounted for by this reason.

B. Lack of knowledge about the law and/or guidance

Very few of the missing trials in the Compliance Evaluation Program I can be explained by lack of knowledge of the law because FDA sent each sponsor a letter specific to the protocol that included information about the law and guidance. FDA mailed 1748 letters to sponsors of protocols submitted to CDER between January 1, 2002 and September 30, 2002. The difference (324) between the total protocols reviewed (2062) and the number of letters sent (1748) is explained by the fact that multiple protocols were included in one submission to FDA and no letters were sent to NIH. We did not follow-up with sponsors via telephone in Compliance Evaluation Program I as we did in Compliance Evaluation Program II due to the volume of calls that would have been required and the decline in project resources. However, we know from the review of 2004 oncology data that only 18% (2/11) of trials not listed could be accounted for by this reason.

C. Identification of trials to test effectiveness

Although the definition of "effectiveness" was not mentioned in the follow-up telephone calls, we still consider it a possible reason why sponsors are not listing some of their trials. In this study, we considered trials with either primary or secondary efficacy endpoints as trials to test effectiveness.

An example of a possible inconsistency in identifying effectiveness is the following: A sponsor submitted a phase 2 study to evaluate the safety, tolerability and antiviral activity of an investigational drug in HIV-infected children. The primary endpoint was to evaluate the safety and tolerability of the drug. The secondary endpoint was to evaluate the antiviral activity of the drug. FDA identified this trial as a trial to test effectiveness and considered it to have met the guidance criteria for inclusion, however, the sponsor did not list the trial in ClinicalTrials.gov.

For Compliance Evaluation Program I, we did not record how many protocols included primary versus secondary efficacy endpoints and thus cannot estimate the number of missing protocols that could be accounted for by this data element. We modified our database in Compliance Evaluation Program II to include primary and secondary efficacy endpoints. Of the protocols meeting the criteria to be listed in ClinicalTrials.gov (n = 55), 38 had both primary and secondary efficacy endpoints, 13 had primary efficacy endpoints, and four had only secondary efficacy endpoints. Of the four protocols containing only secondary efficacy endpoints, two were listed in ClinicalTrials.gov. For the two trials not in ClinicalTrials.gov, discussions with pharmaceutical company representatives did not suggest the definition of effectiveness as the reason for not listing the trial.

D. Concerns about availability of information to the public and to competitors

We suspect some pharmaceutical companies decided not to list their trials in ClinicalTrials.gov or to list limited information about their trials due to concerns about competitors gaining insight to their drug development plan. We know from the review of 2004 oncology data that 9% (1/11) of trials not listed could be accounted for by this reason. However, the statutory provision mandating the establishment of the data bank does not contain an exception from the requirements for information sponsors consider proprietary. The only ground in either FDAMA 113 or the FDA guidance for not including information about an investigation in the data bank is if a sponsor provides a detailed certification to the Secretary of HHS that disclosure of such information would substantially interfere with timely enrollment of subjects in the clinical trial. FDA has not identified specific instances when disclosure of information would substantially interfere with enrollment of subjects in a clinical investigation. In the guidance to industry, we solicited comments on this topic for the purpose of including a listing of acceptable reasons for certification. We received no comments and have received no certifications from sponsors.

The issue of making information publicly available raised questions about the validity of this concern and its impact on sponsors' listing of trials to the Clinical Trials Data Bank. To better understand the scope of this concern, we explore some of these issues in the next chapter.

 CHAPTER 5: PUBLIC AVAILABILITY OF INFORMATION

In this section, we reviewed the types of information publicly available about the trials that were either not listed in ClinicalTrials.gov or were listed with limited information about the location, drug name, or sponsor name.

I. Trials Not Listed in ClinicalTrials.gov

A. Objective

The objective of this review was to determine what information about drugs in the pipeline is available on the Pharmaceutical Research and Manufacturers of America (PhRMA) website and to compare that information to the information available on ClinicalTrials.gov. PhRMA, a trade organization representing research-based pharmaceutical and biotechnology companies, hosts a New Medicines in Development website at http://www.phrma.org/newmedicines/. According to PhRMA, this website includes treatments under development by and in the pipelines of America’s biopharmaceutical companies.

B. Methods

  • We conducted reviews at five different times between August 2003 and March 2005: August 2003, January 2004, April 2004, May 2004, and March 2005.12
  • These reviews captured the following information posted on the PhRMA website and the ClinicalTrials.gov website:
    • drug name
    • sponsor
    • phase
    • trial listing in PhRMA
    • trial listing in ClinicalTrials.gov
  • The PhRMA.org website was reviewed to identify drugs under development in phase 2, 3, or 4 for breast cancer, brain cancer, stroke, Parkinson's Disease, and Alzheimer's Disease. PhRMA's website does not list specific trials.
  • If the drug listed by PhRMA was located in ClinicalTrials.gov as an open or closed study in any phase and conducted by the company or another sponsor (e.g. NIH), then it was considered a trial listed in ClinicalTrials.gov.
  • The results were tabulated for comparison.

C. Results

Of the five diseases reviewed, pharmaceutical companies listed information about their drugs on PhRMA.org more often than on ClinicalTrials.gov. However, the 2005 review showed an increase in trial listings on ClinicalTrials.gov over those listed in 2003 and 2004 in three of the five diseases (brain cancer, Parkinson's Disease, and Alzheimer's Disease). Table 8 shows the number of drugs listed on PhRMA.org as compared to ClinicalTrials.gov.

 Table 8. Drugs Listed in PhRMA.org and ClinicalTrials.gov
  2003 2004 2005
Disease Drugs listed in PhRMA.org(phase 2-4) Drugs listed in ClinicalTrials.gov Drugs listed in PhRMA.org
(phase 2-4)
Drugs listed in ClinicalTrials.gov Drugs listed in PhRMA.org
(phase 2-4)
Drugs listed in ClinicalTrials.gov
Breast Cancer 30 23 (77%) 27 19 (70%) 21 15 (71%)
Brain Cancer 12 7 (58%) 9 6 (67%) 6 6 (100%)
Stroke N/A N/A 11 6 (55%) 10 5 (50%)
Parkinson's Disease N/A N/A 17 7 (41%) 13 11 (88%)
Alzheimer's Disease N/A N/A 15 5(33%) 15 10 (67%)
N/A: Diseases were not reviewed during the specified year.

 

D. Discussion

It is possible that a trial for the drug listed in PhRMA.org existed in ClinicalTrials.gov but the investigator was unable to find it because the sponsor chose not to list the specific drug name in ClinicalTrials.gov (see Section IIB below).

II. Trials Listed in ClinicalTrials.gov with Limited Information about Location and Contact Information, and Drug and Sponsor Name

There are three data fields for which some companies provided limited information for listed trials. These fields are location, drug name, and/or sponsor name. We will discuss the review of each field separately below.

A. Location and Contact Information

1. Objective

The objective of this review was to determine what types of information sponsors submitted to ClinicalTrials.gov to describe the study location and contact information.

2. Methods

  • Between July 2002 and August 2004 FDA sent 23 letters to sponsors via e-mail referencing a total of 68 trials listed in ClinicalTrials.gov that did not list trial site locations (see appendix P for a copy of the letter sent to sponsors). In September 2004, NLM assumed the role of sending the letters electronically.
  • In the letter, specific trials were referenced, and sponsors were asked to include the city, state, and country for each clinical trial site so that visitors to ClinicalTrials.gov could search for clinical trials by location.
  • We retrospectively sampled records four times during the following periods to review location and contact information: October 2003, September and December 2004, and March 2005.
  • In October 2003, we randomly sampled 100 records listed in ClinicalTrials.gov between August 2001 and September 2003.
  • We reviewed all new trial records entered in ClinicalTrials.gov for the following months:
    • May 2004 (n = 33)
    • August 2004 (n = 64)
    • January 2005 (n = 53)
  • We determined if the sponsor listed the following location information:
    • Institution where the trial was being conducted
    • City and state where the trial was being conducted
    • Principal investigator was identified
  • We determined if the sponsor listed the following contact information:
    • Local telephone number
    • Central telephone number (800 number for a clinical trial call center)
    • Both a local and central contact number
    • Did not list a telephone number

3. Results

Table 9 shows greater than 90 percent of sponsors listed the trial site's city and state, but less than half listed the name of the institution and principal investigator. Approximately one-half to two-thirds of sponsors chose to list a central contact telephone number rather than a local telephone number.

 Table 9. Location and Contact Information Listed by Pharmaceutical Companies
Data Element Percent of Trials Listed
October 2003(n = 100)
Percent of Trials Listed
May 2004
(n = 33)
Percent of Trials Listed
August 2004
(n = 64)
Percent of Trials Listed
January 2005
(n = 53)
LOCATION        
Institution 64% 73% 40% 42%
City and State 93% 100% 95% 91%
Principal Investigator 45% 48% 30% 30%
CONTACT
NUMBER
       
Local 40% 27% 33% 36%
Central N/A 42% 54% 64%
Both N/A 12% 5% 13%
None N/A 18% 8% 13%
N/A: Investigator reviewed trials for local telephone number only.

4. Discussion

In accordance with Section 113 of FDAMA, sponsors are required to list the location of the trial and contact information for all trial sites listed in ClinicalTrials.gov. Location of the trial is important information for patients. Some patients are willing and able to travel anywhere to participate in a study, but others are unwilling or unable to travel. Those in the latter group can limit their searches to locations to which they can travel. Initially, sponsors were asked to include in the trial record the name of the institution where the clinical trial was to be conducted. Some companies chose to list the city and state but not the institution name. Some companies chose to list an 800 telephone number only without listing any location information.

The PRS data fields were modified to reflect that an adequate description of the trial site location would consist of the city and state for each trial site and recognize that from a patient perspective, one cannot discern the trial site location from an 800 number. For example, if a patient conducted a search for asthma trials and 40 trials appeared with only 800 numbers listed and no location, the patient would possibly have to make 40 individual phone calls to determine each trial's location.

B. Drug Name and Sponsor Name

1. Objectives

The objectives of this section were as follows:

  • To determine to what extent sponsors listed their company names or drug names for trials listed in the Clinical Trials Data Bank.
  • To determine what information was publicly available for trials without drug names listed in the Clinical Trials Data Bank.

2. Methods

1) To determine the extent of company and drug names listed in ClinicalTrials.gov, we reviewed all protocols submitted by pharmaceutical companies between May 2001 and February 2004 (n=1220). We identified trials whose sponsors listed "investigational drug" instead of the drug name and "pharmaceutical company" instead of the actual company name.

2) To determine the availability of public information for trials without drug names listed, we conducted two reviews in 2002 and 2004 as follows:13

  • In 2002, we reviewed the names for seven drugs listed as "Investigational Drug" in ClinicalTrials.gov.
  • In 2004, we reviewed the names of twelve drugs listed as "Investigational Drug" in ClinicalTrials.gov, of which six were FDA approved drugs and six were investigational drugs.
  • Searches were conducted using the drug name on the company website, PhRMA.org, Google.com, and competitive intelligence websites.

3. Results

In the 2002 search for publicly available information about trials listed without drug names, six of seven drugs were found on the companies' websites. Most of the information was found in the annual report or investor materials section of company websites. More detailed information could be obtained from other websites for a fee.

In the 2004 search, results were similar to those in 2002. Each of the twelve drugs was identified in at least one of the sites mentioned above.

In the February 2004 review to evaluate to what extent sponsors listed their company names or drug names in ClinicalTrials.gov, we found that five pharmaceutical companies did not list their company name and nine companies did not list their drug name in ClinicalTrials.gov. Further analysis showed the following:

  • 112 out of 1,220 (9%) trials did not contain drug name (e.g. "Investigational New Drug" used instead).
  • 96 out of 1,220 (8%) trials did not contain company name (e.g. "Pharmaceutical Company" used instead).

Of the trials mentioned above, 80 omitted both company name and drug name.

4. Discussion

Some sponsors did not list their company name and/or drug name for some or all trials. Instead they listed "Pharmaceutical Company" or "Investigational New Drug" in the sponsor name and drug name data fields.

Information about the drug name and company name could be important to patients. For example, if a patient learns about a new cancer treatment on the evening news, she might remember the name of the drug or the name of the company investigating the drug. In either situation, if the patient searched by the drug name or company name in ClinicalTrials.gov, she would not find it if the sponsor had not provided this information.

Based on the results of the reviews, it is evident that even if a drug's name is not disclosed on ClinicalTrials.gov, it generally is available on publicly-accessible websites.

Patients are the most likely population to benefit from having the drug and company names available to them when searching ClinicalTrials.gov. Competing companies and investors already have access to extensive information about investigational drugs through general Internet searching and purchasing information from subscription databases such as Biospace's Clinical Competitive Intelligence System (CCIS) or for a one-time fee-per-information packet. Futhermore, it is consistent with the intent of ClinicalTrials.gov for this information to be available to patients in one location, rather than their having to search multiple websites and attempt to collate disparate information.

In the fall of 2004 it became evident that sponsors were modifying their practices with regard to listing drug and sponsor names in the data bank; by December 31, 2004, the major pharmaceutical companies were listing their company names, and it appeared that only one major pharmaceutical company routinely continued to not list drug names. It also appears that some companies continue to make case by case decisions to not list drug names.

C. 2004 Update

The National Library of Medicine provides us with a list of trials submitted to ClinicalTrials.gov each week by non-federal sponsors. From the list we are able to monitor the trend in the number of trials submitted on a monthly basis.

Figure 6 shows that in the first two years after implementation (June 2002 through May 2004), an average of 44 trials/month were listed. Since June 2004, an average of 71 trials per month were listed. In the summer of 2004, media and congressional attention to issues related to the public availability of clinical trial information may have heightened awareness about ClinicalTrials.gov.

 Figure 6. Monthly Trial Listings in ClinicalTrials.gov by Non-Federal IND sponsors 
Figure 6. Monthly Trial Listings in ClinicalTrials.gov by Non-Federal IND sponsors

Similarly, the results from Compliance Evaluation Program II compared to the results from Compliance Evaluation Program I showed about a twenty percent increase in compliance (50% to 69%) for non-federal sponsors.

We hope that compliance will continue to increase. In addition to continued media and congressional attention to this topic, on January 6, 2005, PhRMA issued a press release about its new voluntary disclosure policy, "Pharmaceutical Companies to Make More Information Available About Clinical Trials." Under this policy PhRMA member companies should provide information about clinical trials for all diseases, not only for serious or life-threatening diseases. The policy states that information about new hypothesis-testing trials will be posted on ClinicalTrials.gov on a voluntary basis beginning July 1, 2005, and ongoing hypothesis-testing trials are to be posted by September 13, 2005.

 CHAPTER 6. REPORTING CLINICAL TRIAL RESULTS IN THE CLINICAL TRIALS DATA BANK

Recent public attention has focused on expanding public access to information about results of clinical trials. Proposals for establishing results databases have been offered and/or implemented by a variety of organizations, such as the American Medical Association,14 National Institutes of Health,15 International Committee of Medical Journal Editors,16 the US Congress, 17 the World Health Organization,18 PhRMA,19, 20, 21 and individual pharmaceutical companies.22,23 FDAMA 113 provides for including information pertaining to the results of clinical trials, with the consent of the sponsor.
This chapter explores issues around clinical trial results and other information submitted by sponsors to the Clinical Trials Data Bank.

Background

Currently, there is no Congressionally-mandated data bank for results of clinical trials. Listing results of clinical trials in ClinicalTrials.gov is voluntary. In order to provide sponsors with a mechanism to reference results, ClinicalTrials.gov created a general More Information (formerly called Related Information) data element field. Sponsors can provide References, defined as citations to publications related to the protocol such as background and/or results. Sponsors provide either unique PubMed Identifier (PMID) of an article or enter the full bibliographic citation.

In October 2004, ClinicalTrials.gov was updated to include a specific Results Reference data element field to allow sponsors to indicate if the reference provided reports on results from the referenced clinical research study.

The More Information data element field also includes an option to provide other Links, defined as a Web site directly relevant to the protocol. It specifies that links should not include sites whose primary goal is to advertise or sell commercial products or services. Links to educational, research, government, and other non-profit Web pages are acceptable. All submitted links are subject to review by ClinicalTrials.gov.

I. Objectives

The objectives of this review were to determine the following:

  • Which citations to publications related to the protocols sponsors listed in ClinicalTrials.gov
  • To what extent publication citations (referred to as References) were available at no cost and as full-text
  • What types of links to websites did sponsors list in ClinicalTrials.gov
  • To what extent the links listed by sponsors contained promotional material.
II. Methods

A. References

NLM staff identified all records submitted through the PRS by non-federal sponsors through October 2004 (n=1768).

  • We reviewed each record via searches of the publicly accessible ClinicalTrials.gov website to determine the type of information being posted and if the information was available as no-cost, full-text articles.
  • Information was categorized into 14 categories (see Table 10).
  • Data were entered in a Microsoft Excel spreadsheet. References were further evaluated through PubMed abstracts. For several references, their category was defined by PubMed and listed at the bottom of the PubMed abstract. References without a PubMed category were reviewed and categorized.
  • The references were reviewed to determine the availability of the information as no-cost, full-text articles.
 Table 10. Categories of References
Clinical Trials Animal Studies
Human Studies Review Articles
Case Reports Letters
Editorials Cellular/Biological Studies
Evaluation Studies Textbooks
Data Reports Practice Guidelines
Informative Articles Other (not yet classified)

B. Links

  • NLM staff identified all records submitted through the protocol registration system (PRS) by non-federal sponsors through October 2004 (n=1768).
  • We reviewed each record on ClinicalTrials.gov to assess what types of links were being listed and to assess whether the links contained promotional material.
  • The links were reviewed for promotional content. Only studies that were currently recruiting were evaluated (n = 111).
III. Results

A. References

Seventy-six records (4%) were identified as having posted references in the More Information data field. Of the 76 records, 66 were for unique interventions (drug, device, or biologic). The 76 records contained 358 references and were listed by 44 unique non-federal sponsors. Of the 358 references reviewed, 34.6% described human drug studies. Other commonly referenced materials included animal studies (21.8%), review articles (11.7%), and human studies not using drugs (11.2%). The remaining 20.7% of the references cited a variety of other types of information such as editorials, case reports, correspondence, textbooks, and practice guidelines. Of the 358 references reviewed, 177 (49%) were available as full-text. Of these, half (85) were available at no cost. Many important medical journals make available online, free and full text articles and release some or all of their content 6-12 months after publication. We did not identify the time at which free access was available for each journal publication, nor did we do a search to determine which completed studies had results published.

B. Links

One-hundred records were identified as having posted links in the More Information data field. The 100 records contained 139 links consisting of 78 unique sites and 61 duplicates. Approximately half (51%) of the links were to sponsors' websites. The remaining links were to clinical trial search engines (18%), trial sites (14%), nonprofit sites (9%), government sites (4%), drug sites (3%), and research facility sites (1%). A preliminary review by OSHI found none of the links contained promotional material. We did not have access to information about these products sufficient to assess whether information in these sites was selectively presented or otherwise misleading with respect to evidence of safety or efficacy of products discussed. However, we found no links to information that appeared overtly promotional, necessitating referral to FDA/CDER Division of Drug Marketing, Advertising and Communication or the CBER Office of Compliance and Biologics Quality.

 CHAPTER 7. LIMITATIONS

The limitations of Evaluation Programs I and II are listed below:

Compliance Evaluation Program I

  • The study reviewed protocol listings for applications submitted to CDER during a nine-month period in 2002. Although no CBER protocols were reviewed, we expect the results for CBER applications would be similar.
  • As part of Compliance Evaluation Program I, FDA did not contact IND sponsors to inquire why protocols that FDA determined to meet the criteria for inclusion were not included in ClinicalTrials.gov. This was due to the volume of calls that would have been required and the decline in resources.
  • FDA does not maintain a list of serious diseases and conditions to assist sponsors in deciding if a protocol should be listed in ClinicalTrials.gov. It was evident from discussions during this project that the seriousness of a disease is often a matter of judgment and can vary by protocol. For example, the seriousness of angina can vary greatly depending on whether it is stable or unstable, new-onset or chronic. In order to ensure consistency in how we coded each protocol, it was necessary to create a list of serious and non-serious diseases and conditions. This list does not have official status for any purposes other than this project.
    • There were eight diseases that we considered serious for purposes of this project that could have been considered non-serious. We analyzed the data using both classifications. If we would have considered the eight diseases non-serious, sponsor participation would have increased from 33% to 37%.
  • We were unable to retrieve 24 paper IND submissions. We do not expect these missing documents would impact the overall study results.
  • We were unable to identify the phase in six percent (127/2,062) of the protocols submitted to CDER. If the phase was not specified, we did not consider it a trial required to be listed in ClinicalTrials.gov even if it was for a serious disease and tested effectiveness.
  • Due to resource limitations we limited the timeframe for data collection to nine months. We do not expect that three additional months of data would impact the overall study results for Compliance Evaluation Program I.

Compliance Evaluation Program II

  • Due to resource limitations we limited the timeframe for data collection to three months. We selected the timeframe of May-July in order to reflect the potential increase in trial listings due to the heightened awareness about ClinicalTrials.gov. Because of increased attention to issues related to the public availability of clinical trial information, it is possible, but not likely, that overall compliance for twelve months would have increased had we reviewed twelve months of data instead of three months.

Compliance Evaluation Programs I and II

  • The studies reviewed only commercial protocols submitted to FDA during the study timeframes. We recognize that protocols coded research may have met the criteria for inclusion in ClinicalTrials.gov; however, due to limited resources we chose to review only commercial protocols.
  • We coded trial phase by using the protocol synopsis or the protocol title. If phase was not listed in either of these sections, we used the phase listed by the sponsor on the FDA Form 1571. It is possible that these phase listings were not always correct. We estimate that phase was identified using Form 1571 in approximately 10% of the protocols.
  • Some draft protocols were coded as new protocols (PN) in COMIS. We did not include these protocols in the studies. However, it is possible that some protocols that were submitted as draft protocols were not easily identified as a draft. These protocols would not have been required to be listed in ClinicalTrials.gov.
  • Multiple steps were taken to verify trial listings in ClinicalTrials.gov. It is possible that a listing could have been missed if serial numbers in the OSHI database and ClinicalTrials.gov didn't match and:
    • Sponsors used different protocol numbers for FDA submissions and ClinicalTrials.gov submissions
    • Final versions of the protocols submitted after September 30, 2002 were not identified.

 CHAPTER 8. SUMMARY

The Department of Health and Human Services developed ClinicalTrials.gov in response to legislation calling for a publicly-accessible registry of clinical trials for serious or life-threatening diseases and conditions. Compliance with the legislation has been mixed. While progress has been made, participation by the pharmaceutical industry is less than expected despite a federal law, a final guidance document, a targeted education program, and an easy-to-use web-based data entry tool. Some pharmaceutical companies do not provide required clinical trials, some provide only limited information, while others voluntarily list trials that go beyond the criteria specified in the guidance.

There has been progress on implementing the legislation, however more needs to be done by FDA, pharmaceutical companies, and others to assure increased participation in ClinicalTrials.gov.

  • FDA should further clarify messages about which trials and what information should be listed in ClinicalTrials.gov. We have updated IND acknowledgement letter templates to include a new paragraph reminding sponsors of their responsibility to comply with Section 113 of FDAMA and encouraging the listing of all trials. We will update the Guidance for Industry Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions to reflect the findings from our study.
  • Pharmaceutical companies and other private sector sponsors are encouraged to review their systems for identifying and submitting protocols to ClinicalTrials.gov. We expect that the number of industry-sponsored trials submitted to ClinicalTrials.gov over the next six months will continue to rise as a result of a recent PhRMA initiative and public awareness and scrutiny. Under a new voluntary disclosure policy announced in January 2005, PhRMA members have agreed to provide information about ongoing hypothesis-testing trials for all diseases to ClinicalTrials.gov by September 13, 2005. The policy also encourages PhRMA member companies to establish and make public procedures for verifying compliance with the policy. In addition, the ICMJE announced their position that trials must be registered in order to be considered for publication. New trials will be required to be registered starting July 1, 2005 and ongoing trials must be registered by September 13, 2005.
  • Patient advocacy groups should continue to be proactive in encouraging FDA and pharmaceutical companies to make information about ongoing trials more available through ClinicalTrials.gov. Advocacy group initiatives like the Kidney Cancer Association's policy to not list a clinical trial on its website unless the trial is listed in ClinicalTrials.gov are commendable.

The collection and dissemination of information about clinical trials and their outcomes is an important consumer and health practitioner issue. FDA will continue to encourage sponsors to put required and voluntary information into ClinicalTrials.gov. We believe a comprehensive clinical trials database can lead to more efficient and timely discovery of the answers to scientific questions that will result in more quickly learning about the safety and efficacy of treatments for patients.

FDA welcomes a continued dialogue on the type of information from clinical trials that would be useful to patients, families, and providers to facilitate their making better informed treatment decisions. Comments can be submitted via email to the Office of Special Health Issues at 113trials@oc.fda.gov or by mail to Director, Office of Special Health Issues, Office of External Relations, Office of the Commissioner, HF-12, 5600 Fishers Lane, Rockville, MD 20857.


 1 Final Guidance for Industry: Information Program on Clinical Trials for Serious or Life-Threatening Diseases and Conditions [PDF: http://www.fda.gov/OHRMS/DOCKETS/98fr/00d-1033_gdl0003.pdf] Because the Agency does not approve protocols, we have interpreted this to mean within 21 days after the trial is open for enrollment.
2See http://www.fda.gov/OHRMS/DOCKETS/98fr/001033gl.pdf
3See http://www.fda.gov/OHRMS/DOCKETS/98fr/001033gd.pdf
4See http://www.fda.gov/cder/guidance/4856fnl.htm
5See http://www.fda.gov/OHRMS/DOCKETS/98fr/2004d-0014-gdl0001.pdf
6 See http://www.fda.gov/cder/guidance/3082fnl.pdf When FDA receives an IND it is categorized as either "commercial" or "research." A commercial IND is one in which the sponsor is a corporate entity (rarely, some other organization seeking to develop a drug for marketing); a research IND sponsor is typically an individual investigator, academic institution, or the NIH. FDA may designate an IND as commercial if it is clear the sponsor intends the product to be distributed in interstate commerce at a later date.
7http://www.fda.gov/cder/guidance/5645fnl.htm
8http://www.fda.gov/cder/guidance/5645fnl.htm
9Although most of the fast track drugs that have been approved since 1998 have been for cancer or AIDS-related indications, there are many fast track designations that have been granted for a variety of serious diseases. As of December 31, 2004, CDER received 328 fast track designation (FTD) requests. Of these, 77 were denied and 13 were pending action. Of the 238 FTD requests granted, 49% (116/238) were for cancer or AIDS-related indications. The remaining 122 FTD requests granted were for other serious diseases, e.g. Parkinson's Disease, Amyotrophic Lateral Sclerosis, Huntington's Disease, Multiple Sclerosis, Chronic Spinal Cord Injury, Acute Stroke, Acute Pancreatitis, Sickle Cell Disease, Obesity, Complicated Skin Infections, Mucositis, Macular Degeneration, Systemic Lupus Erythematosus, Rheumatoid Arthritis, Osteoarthritis, Cystic Fibrosis, Acute Bacterial Infections, Candidiasis, Aspergillosis, Pneumocystosis, and Congestive Heart Failure. OSHI data on file.
10 Cancer protocols are submitted to multiple divisions in CDER. The study was limited to protocols submitted to CDER's Division of Oncology Drug Products, HFD-150.
11http://www.cfsan.fda.gov/~frf/forum03/U-04.htm
12OSHI data on file
13OSHI data on file

The hyperlinks in the footnotes below will take you off the FDA Web site: 
14June 2004 http://www.ama-assn.org/ama/pub/category/14314.html
15September 2004 http://www.publicaccess.nih.gov/
16September 2004 ICMJE announces clinical trials must be listed in a public trials registry to be considered for publication http://www.icmje.org/clin_trial.pdf and May 2005 http://www.icmje.org/clin_trialup.htm.
17October 2004 Kennedy/Dodd Bill S 2933 Draft and Markey/Waxman F.A.C.T. Bill HR 5252 Draft
18October 2003, November 2004 http://www.who.int/rpc/meetings/en/WHO2.pdf and April 2005 WHO technical consultation on clinical trial registration standards meeting http://www.who.int/ictrp/news/ictrp_sag_meeting_april2005_conclusions.pdf
19June 2004 PhRMA updates principles for communication of clinical trial results and conduct of clinical trials http://www.phrma.org/publications/publications//2004-06-30.1035.pdf
20September 2004 PhRMA announcement of central database presenting results of clinical studies of marketed drugs.
http://www.phrma.org/mediaroom/press/releases/07.09.2004.1063.cfm
21January 2005 Joint Position on the Disclosure of Clinical Trial Information via Clinical Trial Registries and Databases and related information. http://www.phrma.org/mediaroom/press/releases/06.01.2005.1112.cfm and
http://www.phrma.org/mediaroom/press/releases/06.01.2005.1114.cfm
22December 2004 http://www.lillytrials.com/
23September 2004 http://ctr.gsk.co.uk/welcome.asp


APPENDICES

Appendix A: Acronyms and Definitions

Appendix B: CDER Letter Template

Appendix C: CDER Education Program

Appendix D: CBER Letter Template

Appendix E: CBER Education Program

Appendix F: Compliance Evaluation Program I: Data Collection Process

Appendix G: Protocol Form 2

Appendix H: List of Serious and Non-Serious Diseases and Conditions

Appendix I: Compliance Evaluation Program I: Additional Data Elements

Appendix J: Data Elements Extracted for Office of Women's Health Project

Appendix K: Office of Women's Health Data Entry Form

Appendix L: Letter Results Data Entry Form

Appendix M: Thank You Letter

Appendix N: Compliance Evaluation Program II: Data Collection Process

Appendix O: Submissions Form

Appendix P: Trial Site Location Letter

Appendix Q: Document Check-In Data Entry Form

 

horizontal rule