Clinical endpoints in diabetes studies: a new approach

Jonathan Goldman, M.D. • FACC Chief Medical Officer, ICON Medical Imaging

Didier Saur, M.D. • Vice President and Global Therapeutic Group Leader, ICON Clinical Research

Diabetes Studies – Changing Objectives?

The pressure on national health care systems to have more effective drugs available to treat diabetes and prevent its complications, is well documented. All the anti-diabetic drugs approved by the Food and Drug Administration (FDA), European Medicines Agency (EMEA) or other Regulatory Authorities have been indicated to improve glycemic control, with glycosylated hemoglobin (HbA1c) being seen as the “gold standard” surrogate endpoint for the approval of products.

However, questions are now being raised regarding the core objective of clinical drug development in diabetes. Is it to decrease HbA1c, or is it to decrease the risk of occurrence of the most serious complications of diabetes, such as death, myocardial infarction and stroke? Recent publication of the results of studies such as ACCORD, ADVANCE or VA-DT are focusing the minds of both clinicians and drug developers.¹ Data from newer drugs designed to treat type 2 diabetes have shown that optimal control of glucose levels in diabetic patients has not resulted in a reduction in the cardiovascular risk to patients. The drug classes of potential concern for a direct cardiotoxic effect include GLP-I analogues, DPP-4 inhibitors and thiazolinediones. In fact, cardiovascular risk may even be increased in some cases.²

The fact that marketed anti-diabetic drugs may have worsened the cardiovascular risk despite adequate treatment of diabetes, illustrates starkly how safety and efficacy are intimately linked in metabolic disorders. Adequate glucose control is required to prevent acute metabolic complications, and less serious effects of this condition, such as eye disease. As a result, FDA requirements for demonstration of the safety of anti-diabetic medications is increasingly designed to exclude increased cardiovascular risk rather than to demonstrate the beneficial effect of the treatment on the cardiovascular risk.

Evolving FDA Guidelines – A Need for Adjudication

In July 2008 the FDA convened an Endocrine and Metabolic Drug Advisory Committee (EMDAC) to consider the issue of cardiovascular risk in the development of drugs for the treatment of type 2 diabetes. This Committee agreed it would be appropriate to conduct clinical trials in Phase II to III to rule out an unacceptable risk, rather than be required to demonstrate a cardiovascular benefit. In addition the EMDAC voted overwhelmingly that anti-diabetic drugs without safety signals should still be required to undergo a long-term cardiovascular trial.3 These requirements would also apply to marketed products.

Building on the recommendation of the Advisory Committee, in December 2008 the FDA published their guidance document on the evaluation of cardiovascular risk in new therapies for type 2 diabetes.⁴ While this document does not bring significant new elements compared to the earlier EDMAC recommendations, it does document requirements for both scope and methodology. This guidance document represents only the current views of the FDA, but it is likely that other agencies such as the EMEA will have similar requirements.

The implications of this FDA guidance on current and future diabetes studies is clear—Cardiovascular (CV) events should be collected, documented, and adjudicated by independent experts.

The enrolment criteria for the study, and the study duration, will dictate the number of investigator reported CV events. Obviously the event rate is also very dependent on the exact definition of CV events. For the FDA the composite CV endpoints requiring adjudication should include cardiac death, acute myocardial infarction and stroke. However, the requirements and definition of CV events will potentially evolve and ICON would recommend the expansion of the definition of the composite CV endpoint to also incorporate acute coronary syndromes, and revascularization procedures (including CABG, percutaneous coronary interventions (PCI), and interventions for peripheral vascular diseases). Depending on the enrolment criteria it may also be beneficial to include hospitalization for chronic heart failure and transient ischemic attack (TIA) in the composite CV endpoint.

Clearly, a wide variety of clinical data will need to be collected to correctly adjudicate each of these complex endpoints. It is also generally believed that investigators tend to over-diagnose and over-report possible CV events in these categories, and do not use standardized criteria. For this reason a centralized system is required with independent adjudicator experts.

Traditional Independent Endpoint Adjudication— An Inefficient Process

Many clinical trials incorporate the evaluation of endpoints for efficacy and/or safety by independent adjudication committees, usually referred to as Clinical Event Adjudication Committees (CEC). The CEC are often comprised of physicians with varied expertise in the relevant therapeutic areas and other diagnostic areas. The CEC reviews all relevant clinical and diagnostic data to provide an independent, blinded adjudication of trial endpoints or events. Historically, most adjudication work has been performed in the academic sector, in a paper-based environment that relies on the evaluation of paper CRFs, listings, videotapes, or films, and does not allow simultaneous project tracking, nor provide an optimal audit environment. Despite high-level physician expertise, the conduct of the committee can be impacted by lack of real-time access to all relevant data in an electronic manner and the ability to easily identify multiple data.

Utilising Technology to Enhance the Adjudication Process

Given the inefficiencies inherent in the current adjudication process, finding a way to give committee members real-time access to data would significantly enhance and expedite the process. ICON has found a way to do this utilising technology developed by its Medical Imaging division. This group has developed a validated endpoint module within its MIRA™ (Medical Imaging Review and Analysis) data management system that is currently being employed in a number of Phase I, II, III, and IV studies. This module is being utilized by a number of adjudication committees whose members can remotely access the system simultaneously from different locations via a web portal, and is linked to an electronic case report form (eCRF) to capture the committee’s findings. The time from the occurrence of the event to the time of adjudication is also accelerated by the ability of the adjudication monitor from the Medical Affairs group to more efficiently track incoming data.

All data relevant to adjudication of potential CV events (for example patient chart, CRF pages, laboratory results, ECG, echocardiogram, CT brain, coronary angiogram, peripheral angiogram, CABG operation note, chest X-rays, computed tomography images) is collected from the sites and appropriately processed for inclusion in the MIRA™ endpoint module. Relevant clinical and laboratory data required to adjudicate events leading to death, primary cardiac and bleeding events are provided through electronic data transfer from the sponsor or their EDC vendor, as well as paper based clinical information. All paper is digitized to create a fully digital environment.

The benefits of facilitating electronic access for committee members are significant. Having remote access to data eliminates the need to bring CEC members together on a regular basis to perform event adjudication and eliminates the need for copying, overnight couriering, and tracking of paper and film data. There is a significant reduction in data loss and the creation of more reliable and secure audit trails. Additionally, programmatic algorithms with logic checks can be incorporated into the eCRF to assist in the adjudication process and committee members can electronically request additional data, allowing the case to remain in the queue of uncompleted cases. Having real-time access to the status of the adjudication process also reduces the risk of over recruitment of patients and the associated study costs.

Quantifying the Cost Savings from Electronic Adjudication

The biggest cost in clinical trial management is per patient investigator fees and ongoing monitoring costs. For event driven studies, there is a critical need to identify when the protocol-defined numbers of endpoints have been adjudicated as definitely present. This can be considered in 3 ways:

1) Is the head-to-head cost of the old versus the new method different in absolute terms? Mathematical modelling data can be used to assess the absolute cost implications of the new approach. For example for a study of 60 months duration and a cost of paper based adjudication of $900 per event package, if the time of the study is shortened by 3 months, then the cost increase for new technology solution has to be less than $0.80 cents per month or $48 per event package that is adjudicated for the course of the study for electronic adjudication to be cost effective (assuming one ignores the additional benefits in terms of efficiency and accuracy described above).

2) Can early identification of target numbers of endpoints reduce clinical trial costs? Early notification of sufficient events can also save clinical trial costs by either reducing the number of enrolled patients or gating duration of follow-up of enrolled patients. Assuming the cost of the older paper and new electronic method is identical, then if a diabetes CV outcome clinical trial incurs fees of $140M ($100M direct fees to clinical research organisation, and $40M indirect fees to Investigator), then for a 60 month study duration, cost savings of $550,000 -$600,000 per week of early closure of enrolment can be anticipated. The savings in direct fees to the CRO could be $400,000 -$450,000 per week. These numbers are all examples and can be refined for specific studies.

3) Can early closure of a study allow earlier or increased revenue from an approved product? Clearly this question depends on the product and its phase of development. Early proof of safety can increase sales by reassurance of prescribing physicians for Phase IV studies. Earlier filing of NDA for non-approved products shortens the time to market, if Phase III studies complete earlier. For products with annual revenues of $100M, this could potentially lead to as much as $2M per month in earlier recognition of sales revenue resulting from earlier completion of a large trial by the use of an electronic adjudication system.

(click Figure 1 below to see an enlarged and more readable version in a new window)

An Outline Data Flow

Figure 1 is an outline of the data flow using ICON’s electronic adjudication system and process. Once a potential endpoint (bleed, stroke, etc.) has been identified at the study site, the investigator will complete an event notification form (ENF) in the EDC system within one business day of identification, providing details of the event and the relevant tests that will be sent within the dossier. This triggers an automatic email notification of the event occurrence to the adjudicator monitor project team at ICON and the data from the ENF will then be uploaded into the ICON’s adjudication system (MIRA) for tracking purposes.

The site should also forward the event adjudication dossier to ICON via traceable courier within two weeks of reporting the event. Once all of the required documents have been received, a QC check on the individual documents in the dossier is performed to ensure that all patient/subject identifiers have been removed and the dossier is loaded into the system along with the patient/subject profile obtained from the EDC system. The adjudication monitor performs a medical review of the dossier within one business day of receipt for medical completeness (to ensure no other potential endpoints or SAEs need to be reported). If there are any issues with the dossier, queries will be forwarded to the site to clarify or obtain any additional documentation as needed. Once the adjudication monitor is satisfied that the necessary documentation has been supplied, and any pertinent queries have been addressed, the dossier is ready for the adjudication committee (CEC). The adjudication monitor will then forward the dossier electronically via MIRA to the CEC for adjudication. During the adjudication process, if the AC requires clarification of any data or document issues, they will raise a query requesting such information within the system. The adjudication monitor will review the queries before directing to the study site or clinical monitors’ attention.

The status of a particular dossier will be tracked throughout the process using MIRA™ enabling the production of reports (e.g. number of events identified but not received by MA, number of events outstanding with CEC, etc.) and query requests. The workflow is demonstrated in Figure 1.

Case Study: Cardiovascular Event Adjudication

A recent large multicenter study required independent adjudication of cardiovascular events for acute coronary syndrome endpoints. Each subject enrolled had a large amount of data collected, received, tracked, randomised and masked at ICON using our MIRA system. This included a clinical profile extracted from the EDC system, other clinical data, serial laboratory measurements, hemodynamic change at stress testing, serial centrally analysed ECGs, centrally analysed coronary angiograms, and centrally analysed rest/stress myocardial perfusion imaging tests. Pre-specified clinical events that had occurred during the follow-up period of the study were also recorded. Some of the data (symptoms, ECG, troponin, myocardial perfusion imaging, and angiogram) was allocated to certain categories such as normal, borderline or abnormal. The raw data, the categories and all images were presented to the CEC in a logic driven eCRF. The logic in the eCRF combined the categories for each subject to yield a programmatically defined diagnosis using a prespecified algorithm.

The CEC could accept, reject or change the categories of any component or the diagnosis overall. If the CEC decided to deviate from the prespecified algorithm, a reason had to be provided. Any discordant cases between CEC members were identified programmatically
and resolved in a consensus meeting.

Conclusion

New regulatory guidance requires independent adjudication of CV events for certain classes of diabetes medication. These studies are large, complex and expensive. Utilising an electronic solution to facilitate event adjudication can bring unique benefits to sponsors, improve study accuracy, efficiency and potentially resulting in significant cost savings.

References:

¹ Skyler JS. Intensive Glycemic Control and the Prevention of Cardiovascular Events: Implications of the ACCORD, ADVANCE, and VA Diabetes Trials. J Am Coll Cardiol 2009;53:298 –304.

² Nissen SE and Wolski K, Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007 356: pp2,457-2,471

³ FDA Endocrinologic and Metabolic Drug Advisory Committee Meeting, 1-2 July 2008, http://www.fda.gov/ohrms/dockets/ ac/cder08.html#EndocrinologicMetabolic

⁴ Guidance for Industry. Diabetes Mellitus – Evaluating Cardiovascular Risk in New Antidiabetic Therapies to Treat Type 2 Diabetes. http://www.fda.gov/cder/guidance/8576fnl.pdf