Docket Management
Docket: 01D-0489 - Draft Guidance: Clinical Trial Sponsors on the Establishment & Operation of Clinical Trial DMCs
Comment Number: EC -11

Accepted - Volume 1

Comment Record
Commentor Dr. Susan Krasny Date/Time 2002-02-19 15:11:31
Organization Stryker Howmedica Osteonics
Category Company

Comments for FDA General
1. General Comments Submitted on February 19, 2002 electronically to: Dockets Management Branch (HFA-305) Food and Drug Administration 5630 Fishers Lane, Rm. 1061 Rockville, MD 20852 Re: Docket No. 01D-0489 Draft “Guidance for Clinical Trial Sponsors on the Establishment and Operation of Clinical Trial Data Monitoring Committees” To Whom It May Concern: Howmedica Osteonics respectfully submits these comments to the above-referenced docket to make sure that the final guidance document is not too prescriptive, particularly for medical device sponsors. Medical device sponsors require guidance that allows the sponsor flexibility in not only determining if a DSMB is required but also in defining their roles and objectives when one is used. We believe there are very unique differences between medical device clinical trials and pharmaceutical clinical trials, particularly when the pharmaceutical trial is government sponsored. The current draft guidance is written with many themes, some of them recurrent. Our comments will focus on the key issues that Howmedica Osteonics observes with this draft, rather than comment on the individual sections of the guidance document. We understand that the designation for this board (e.g., DSMB, DMC, DMB) may change. We have chosen to describe the board in question the Data Safety Monitoring Board (DSMB) throughout this letter. 1. Howmedica Osteonics disagrees with the current document’s position that government sponsors and industry sponsors deal with the same issues regarding DSMBs and therefore the sponsoring organizations do not have to be differentiated. We believe that the basic level of oversight is very different for studies sponsored by these two different groups and therefore the need for a DSMB and the role of the DSMB also differs significantly with the sponsoring organization. Industry sponsors have the additional oversight of the FDA, while government sponsors are the research-supporting institutions with no additional oversight provided. FDA is involved with on-sight monitoring of the IRBs, while NIH trials are limited to an upfront assurance of IRB compliance. 2. Howmedica Osteonics recommends that the “Institutional Review Board Guidebook” published by the Office for Human Research Protections be reviewed/revised together with the current new guidance for Data Monitoring Committees to help ensure consistency across all IRBs and to help ensure there are no gaps in the patient safety oversight provided for each trial. The IRB guidebook provides guidance to IRBs on their jurisdiction and responsibilities, including continuing review and monitoring of the data. We believe that IRB oversight needs to be urgently addressed before the question of whether a DSMB should be added to a trial can be adequately considered. Howmedica Osteonics believes that many IRBs are overwhelmed and confused about their responsibilities when conducting continuing review of clinical trials. Dr. Sugarman, director of the Center for the Study of Medical Ethics and Humanities at Duke University Medical Center, explained that the review boards “have become inundated with (adverse event reports)…Some of the excessive burden that AERs create for IRBs may be attributed to confusing terminology in the regulations that governs trials, differing requirements of the governmental regulatory bodies involved in ensuring patient-subject safety and inconsistencies in the regulations themselves.” The regulations state IRBs must monitor the safety of the patients participating in an IDE trial. Their primary purpose is to assure the protection of the rights and welfare of the human subjects. This is a very large responsibility with limited guidance on the scope of their oversight, consideration of whether they are in a position to carry this out, and how they are expected to do so. 3. Howmedica Osteonics agrees with advice provided by Dr. Janet Wittes during the public meeting held on November 27, 2001 and request that it be added to the current guidance document. She recommended that the sponsor first ask what the safety concerns of the trial are, and then ask how they should be monitored. Additional oversight may not be required by an independent DSMB; it could be in the form of an internal review. The monitoring of patients participating in clinical trials is a multi-layered process that includes the Investigators, the sponsor, and the IRBs. FDA-regulated studies also have the added oversight of the FDA, as well. NIH sponsored trials are required to add a DSMB for all trials involving human subjects. That is not the case for most medical device trials. Once the risk analysis and study design for a new clinical trial has been finalized, a medical device sponsor must determine whether the basic levels of oversight will be adequate for protecting the patients in the trial. The additional oversight provided by a DSMB or other type of committee should be added when there concerns about adequately evaluating accumulating safety data. 4. Howmedica Osteonics believes there are other reasons for a medical device sponsor to consider using a DSMB, other than those described in sections 2.1, 2.2, and 2.3. As a whole, medical device trials are largely unblinded. Depending on the type of investigational device and the comparison treatment or device selected for the trial, even a single-blinded trial may be impossible. The comparison may be another similar device or it may be some other type of standard therapy. In some cases there is no satisfactory prospective comparison available, and a single arm prospective trial design is used. Monitors, sponsors, Film Reviewers, Investigators, and patients may all know what device or treatment each patient in the trial receives. These trials may or may not require additional patient safety oversight, but it may be prudent for the sponsor to consider the use of a DSMB anyway. The oversight of a DSMB in these situations can be extremely useful in advising the trial sponsor during the design and conduct of the trial. Changes that need to be made in the Investigational Plan for a variety of reasons after the trial has be initiated come with much more credibility when recommended by a DSMB. 5. Howmedica Osteonics agrees that the potential risks to trial participants may be a concern if the treatment to be tested is novel, with limited prior information on clinical safety. A DSMB in a new device feasibility study may be required, but their defined role is most likely simpler, such as to advise the sponsor on the ethical design of the protocol and/or to evaluate the final accumulated data for a recommendation for continuing into the next phase of the study. 6. Howmedica Osteonics agrees with section 2.2 regarding the practicality of DSMB review. We suggest that this section be expanded upon to regarding the mechanisms sponsors may need to permit DSMB review. We suggest that if it is not practical due to the speed at which enrollment occurs, but the oversight is necessary, then the sponsor should consider designing enrollment pauses into the enrollment phase of the trial to give the DSMB time to review that data. 7. Howmedica Osteonics disagrees with the current draft guidance definition of an independent DSMB. An overwhelming theme in the current draft, as well as during the public meeting held on November 27th, 2001, was how to prevent the sponsor from seeing the unmasked data. Many medical device trials are unblinded and therefore the sponsor routinely views the trial data. There may be no need for a closed section of the DSMB meeting, as the sponsor personnel most likely prepared the unblinded reports and analyses for the meeting. We disagree with Dr. LeRoy Walters’ statement at the public meeting that if a DSMB is used, the sponsor must have an independent statistician prepare these reports. We agree that the DSMB should still have the option for an “executive” closed session with no participants other than DSMB members. For unblinded studies, we believe that the DSMB is independent as long as there is no sponsor representation on the board itself. 8. Howmedica Osteonics does not believe the DSMB should directly contact the Investigators or their reviewing IRBs. As a medical device sponsor, we would use a DSMB in an advisory role to our company. According to former Secretary Dr. Donna Shalala’s editorial in the NEJM, a new guideline for DSMBs was to be issued that specifies the relationship between DSMBs and IRBs. The current draft guidance document does not address this issue, however. Some have urged for direct communication between the DSMB and the IRBs. Dr. Rick Ferris (NIH/NEI) commented at the November 27, 2001 public meeting that they (NEI) formalize the relationship between the IRBs and the DSMB. NIH guidance published on June 11, 1999 describes the process for the DSMB secretary to send a brief summary report to each investigator after each meeting. The investigator is then required to transmit the report to the local IRB. Howmedica Osteonics does not believe this would be appropriate for medical device trials. We suggest that when feedback is required following DSMB meetings, medical device sponsors prepare a summary for Investigators and their reviewing IRBs, but we do not believe that the DSMB should provide feedback directly to anyone else other than the sponsor representatives. 9. Howmedica Osteonics would like to expand on the definition of the Endpoint Assessment/Adjudication Committee. Adjudication Committees, also known as Clinical Events Committees (CECs), may be used to provide additional oversight during a clinical trial (section 3.3). The current draft guidance document defines the role of this sort of committee. We would like to suggest that an Adjudication Committee is also useful in adjudicating on individual adverse events that are reported (or unreported) in the trial. The committee can review the adverse event reports as submitted by trial investigators and determine if the definitions provided in the protocol are being interpreted in a consistent manner across all institutions. They may determine that certain events that have been reported do not meet the protocol’s definition for an adverse event and should not be reported. They may alter decisions made by the investigator on such things as severity and causality of the event. The committee may provide guidance to the sponsor (and investigators) on adverse event reporting. They may offer opinions or make formal recommendations concerning aspects of the study that impacting a particular patient’s safety. (e.g. safety related protocol changes or input regarding complication rates associated with the subject devices). The committee may be called upon to help the sponsor review potential or actual occurrences of an Unanticipated Adverse Device Effect. If a DSMB is also being used in the trial, some of these reviews may be done by the DSMB in addition to or instead of the Adjudication Committee. 10. Howmedica Osteonics disagrees with the statement that the Adjudication Committee should be masked when performing assessments whether the trial itself is blinded or not. Depending on the medical device and the trial design, the committee may be unmasked when performing these assessments. This may be unavoidable and/or it may be a desirable condition, depending on their defined role. 11. Howmedica Osteonics suggests there is an additional source of external data for the DSMB to review. Medical devices may be simultaneously available for sale outside the US while undergoing an IDE trial in the United States. Device related events occurring outside of the clinical trial, either with the same IDE device or one that is similar, might be reviewed by the DSMB. This review may be done to help the sponsor determine if there is any new adverse information available that would affect the risk analysis for the IDE trial. 12. Howmedica Osteonics suggests there is another consideration to be added to the section on determining the need for a DMC. Many implantable medical device trials are designed to demonstrate the safety and effectiveness of the device within a relatively short amount of time compared to the total life expectancy of the implant. The answers provided in the trial may not satisfy all of the possible long-term safety concerns an FDA advisory panel may have. The FDA may accept a panel’s request to have the sponsor conduct a Post-Approval study or initiate one or more post market surveillance studies. As in the pre-market phase, consideration should be given to whether additional oversight is required for such studies. If so, and a DSMB was already employed during the pre-market phase, it might be reasonable to ask the board to remain in operation during the post-market phase. If so, and a DBM was not already employed, a new one may need to be established. Regardless of any conditions for market approval, the DSMB used in the premarket phase should provide suggestions to the sponsor on the design of future trials for similar devices. 13. While pharmaceutical and NIH sponsors conduct Phase I and II studies, device sponsors conduct feasibility (or pilot) studies. Please add this to the “Early Studies” section 4.4.2. 14. Howmedica Osteonics does not believe it is necessary to add into every future protocol the reasons why DSMB oversight will not be added, when that is the decision that has been made. Dr. William Henderson commented at the public meeting on November 27th that the study protocol should specify that a DSMB was considered, whether one is used or not. We do not routinely conduct clinical trials that would require the additional oversight of a DSMB. Many of the trials are designed to demonstrate equivalency to a comparable marketed device. On the rare occasion that a future trial will use a DSMB, we suggest that only then will it be required to describe the use of a DSMB in the protocol. 15. Howmedica Osteonics agrees that DSMBs should have well-defined procedures, but we do not believe they should have to be routinely provided to the FDA. The protocol will describe the fact that a DSMB will be used and a rough estimate of how often they will meet. 16. Howmedica Osteonics does not think it is necessary to submit DSMB meeting records in the market application to FDA. Additionally, we suggest that the annual progress reports to FDA include the dates the DSMB meetings were held and any decisions that were made at those meetings. Time-sensitive recommendations made by the DSMB to the sponsor will be communicated to the FDA, Investigators, and IRBs, as required. 17. Howmedica Osteonics recommends that clinical trial sponsor consider whether the identity of the board members should be kept confidential, and if so, to whom they should be kept confidential. 18. Howmedica Osteonics does not find the first paragraph in section 2 “Determining Need for DMC” helpful, which discusses when DMCs have generally been established in the past. We suggest that this guidance simply discuss how to determine if a DSMB should be established in future trials. Thank you for the opportunity to provide our input. Respectfully, Susan Krasny, Ph.D. Director, Regulatory Affairs and Clinical Research

EC -11