Data Monitoring Committees: Observations & Issues – Part II
In Part 1 of this post we started the review of some practical issues in the management and functioning of DMCs. We conclude here.
I have seen situations where the sponsor uses the DMC only grudgingly and really does not want the DMC to act as an independent body protecting the patients. The sponsor can be slow in supplying data or fulfilling requests from the DMC or rarely may refuse outright or deny requested information. The sponsor may reject the DMC’s recommendations and cast doubt on the qualifications or integrity of the DMC that they, the sponsor, chose.
Situations like this usually occur when there is a significant safety issue in the study or when administrative problems occur (e.g. errors at study sites, incorrect drug packaging, problems with CROs or monitoring, etc.) that are totally unexpected. Sometimes sponsors are in denial.
This puts the DMC members in a very awkward situation since they want to continue to participate and fulfill their patient protection role but are not allowed to function in the manner they see fit. Resignations from the DMC may thus occur. When this happens, the DMC members who have resigned must make a decision on how to handle their resignations and whom to inform (e.g. FDA, EMA?). This is rare but, alas, does occur.
I have seen multiple charters written which leave out critical elements such as:
- Quorum: How many members must be present at a meeting (whether in person or phone/webinar) for the meeting to be valid? Must the DMC’s findings be consensus and unanimous? What happens if there is a tie vote? For this reason many DMCs have an odd number of members thus preventing a tie vote. What happens if there is not unanimity but a split vote? Is there a minority report?
- Rejection of DMC Recommendations: What happens if the sponsor rejects partially or fully the recommendations of the DMC? Are negotiations possible between the sponsor and DMC? Does the DMC have recourse? Are the DMC opinion and the sponsor refusal reported to FDA and the other concerned health agencies as well as the investigators, IRBs/ECs, partners?
- Blinding/Unblinding: Is this specified in the charter? If blinded, is it also specified when the DMC can ask for and receive unblinded data?
- Mechanism of bringing the decision(s) of the DMC to the sponsor: This is sometimes not adequately specified (letter, minutes, both).
- Minutes: Are they blinded or unblinded or two versions with one blinded the other unblinded? How are the minutes created and issued? When are they issued? If the sponsor has a representative on the DMC to handle the minutes and other administrative activities, is this person blinded or unblinded and how is this handled to protect the study?
- Members: Membership is not always specified? Can an additional or replacement member be appointed and by whom? How is the chairperson chosen?
- Amendments: Can the charter be amended and how? Will it be reviewed yearly (or periodically)?
- Ad hoc: How are ad hoc or emergency meetings handled, particularly if there is an issue that the DMC wishes to discuss without the sponsor representative present?
- Communication: How does the DMC communicate amongst themselves and with the sponsor? Is a sponsor or CRO portal used? Can the DMC communicate outside the portal? Can the DMC meet without a sponsor representative present?
- Use of an outside CRO or company: In many situations, the sponsor may hire a CRO or a company that only handles DMC and operational aspects of clinical trials but not the actual investigator selection, monitoring, etc. The outside vendor will handle all DMC operational aspects and will often supply a statistician taking over these roles from the sponsor. The virtue of this is that it removes the sponsor from having to do these tasks, some of which may require unblinding of the data and the risks of others in the sponsor being inadvertently unblinded. If an outside company is used, there must be a contract between the sponsor and the vendor and the duties should be clearly specified in the contract and, as appropriate, in the charter.The mechanics must also be clear to the DMC members. Whom do they contact with issues, questions, requests? Do they only go through the vendor or are there times when they can or should go to the sponsor. What is the recourse if the vendor or the sponsor is not fulfilling its role adequately?
- The Role of the Chairperson: I have been involved in situations where the sponsor requested that a particular issue or set of data be reviewed and approved or resolved only by the chairperson without the involvement or, sometimes, even the knowledge of the other DMC members. When I was chairperson and such requests were made, I invariably made sure that the other members of the DMC were involved in the issue and that all decisions were made by the full DMC.
Data Presentation, Templates, Shells, etc.
The DMC is usually asked by the sponsor or CRO to prepare or approve the data presentation documents, templates, shells, etc. This can be difficult to do if the DMC does not have a clear idea of the type and quantity of the data being delivered. The DMC rarely if ever sees the case report form or the EDC screens and can only judge the amount and type of data from the protocol(s).
Thus it may become apparent once the data starts arriving or once large quantities of data start arriving that the data presentations must be changed to allow the DMC to do its work adequately.
This is usually not a problem but everyone should be aware that it may occur and it may even be worth writing in the charter that the DMC may make requests for new or different data presentations during the study. The timing for the creation of the new data presentations should also be noted.
In 2012, the DIA Journal published a very germane article: Who Advises the Data Monitoring Committee (DMC)? A Review of Regulatory guidance for Sponsors on DMCs After 5 Years and Advice for DMC Members, Josephine Mary Glover and Richard Kay. DIJ 46(5) 525-531; 2012..
This is an excellent article which points out many tricky and potentially problematic issues with DMCs. It is a “must-read” for any DMC member and covers many areas in depth that are beyond the scope of this posting. One aspect they point out regards liability:
“The DMC members could potentially be held legally liable for their recommendations when patients have suffered harm, either by the sponsor or by patient groups. We would advise potential DMC members to be sure that they have adequate professional indemnity insurance in place. The question, for which no answer is available to our knowledge, is what the level of cover should be. We have encountered a situation in which the sponsor offers to purchase indemnity insurance on behalf of DMC members, which is welcome while not affecting their independence. A preferable contractual alternative would be for sponsors to indemnify DMC members against claims by third parties and to guarantee not to make claims against DMC members themselves, except in case of negligence.” (Page 530).
This is an important aspect of DMCs that has not been discussed much and which may become an issue as DMCs take on expanded roles (e.g. in phase I studies) in more and more studies.
So, in summary, DMCs play a vital role in the clinical trial process and add another level of patient protection. Although they are not quite obligatory, they are clearly best practices and may, de facto, be required by the health agencies before approving new studies. They offer protection not only to the patients but also to the sponsor and everyone else involved in the study. They should be put in place by the sponsor and the sponsor must realize and accept that the DMC may alter the “best laid plans” of the clinical trial. It would be good to see FDA, EMA and other involved health agencies and parties address these issues and put forth new guidelines, or perhaps better, new regulations on DMCs.
Tags: DMCs, drug safety, FDA, healthcare, Pharmacovigilance, regulations