CV Safety for Developers

Regulatory agencies are increasingly concerned about the unwanted cardiovascular effects of new chemical entities and compounds already on the market. Drugs such as Vioxx (rofecoxib), Avandia (rosiglitazone), and Fen-phen (fenfluramine) all have cardiovascular adverse events that were seen in the large prescribing population that were not seen in clinical trials.

Vioxx was withdrawn from the market due to an increase in incidents and thrombotic events, including heart attacks and strokes. Avandia has been shown to lead to myocardial infarction and death of cardiovascular origin. Serotonin receptor antagonists as gastrointestinal therapies have also been associated with cardiac safety issues. Cisapride showed a drug-drug interaction at Cytochrome P450 and induced QT prolongation and torsades de pointes, while Tegaserod is no longer on the market because of drug induced ischemic events.

None of these drugs have a primary target in the cardiovascular system. Drug developers are now being asked to provide evidence of both clinical and nonclinical cardiovascular safety when a compound is systemically available.

It is important that the drug developer maps out a cardiovascular safety strategy in clinical trials. This is done by reviewing the results of pre-clinical data from the guidance delivered by the ICH S7A and S7B in conjunction with results of the clinical pharmacology studies.

Although Phase II and III studies are most often underpowered to detect low frequency cardiac events such as cardiac arrhythmias, a dedicated cardiovascular safety trial is now often required by the FDA. For major drug classes such as antidiabetic compounds and weight loss drugs, longer term effects on the heart need to be assessed. This can include assessing the compound's effect on heart valves and pulmonary hypertension using echocardiography, or looking at the LVEF using echocardiography or a MUGA scan. The class-effect of the compound can be useful in determining what tests are most important.

Antidiabetic therapies

It's difficult to design a "one-size-fits-all" cardiovascular safety strategy for NCEs. We can look at the guidance in one area to see if recommendations can be used in other areas, for example the FDA guidance about new antidiabetic therapies. This guidance suggests a dedicated, randomized, controlled trial for cardiovascular safety that includes:

• A design as an event driven trial to rule out an excess risk measured by an upper bound of a 95% confidence interval for the hazard ratio.

• Enrolling population with a high level event rate is often mandatory.

• Safety committees and adjudication committees to review and adjudicate major adverse cardiovascular events (MACE).

• Ensure that clinical trials (both Phase II and III) designs will allow the conduct of a meta-analysis.

• Risk evaluation and mitigation strategies (REMS) after approval.

It is important to note that these recommendations are, or can be, inter-related.

MACE and its impact on sample size

In efficacy trials, the FDA and EMA recommend that only death, cardiovascular death, myocardial infarction, and stroke be considered as a MACE. However, in safety trials, including hospitalization, cardiac arrhythmias and heart failure episodes would significantly increase the event rate and therefore decrease the sample size. The definition of MACE is critical to compare trials not only within the drug development program, but against other drugs of the same class. In addition to the overall event rate, regulators are concerned about the length and depth of follow-up in large, multicenter trials. The percentage of patients who are "lost to follow-up" can have an impact on the overall conclusion of the clinical trial.

ePRO and outcomes

Electronic patient reported outcomes (ePRO) now plays an important role in clinical monitoring. Although not yet required by agencies, the FDA published a guidance for industry document for patient reported outcomes in 2009.

Patient reported outcomes (PRO) is not a new idea. Usually collected by questionnaires, the ability of these instruments to be put on electronic devices allows efficient capture of this data at various phases of the clinical trial, helping follow the patient between visits. The electronic addition to the PRO can also aid in simplifying the questionnaire. In some cases branching logic can be employed that triggers only on certain responses.

However, the trade-off is to make sure that the instrument provided does not interfere with the correct response. For example, this can be problematic if a "technologically naïve" person cannot operate the device.

REMS

In the FDAAA of 2007, the FDA may require a REMS as part of the drug application. The purpose of the REMS is to ensure the benefits outweigh the risks. With the REMS is a requirement for a medications guide for compounds to be marketed. Medications guides are considered part of the drug labeling and are subject to the same provisions.

Meta analyses

One of the important messages from the new diabetes guidelines is the suggestion of a meta analysis (MA). MAs are a specific statistical procedure where data are combined from multiple trials to make a statement about the overall trend in the compound.

The MA calculates an "effect size" from many studies with similar designs. In studies that look at end-points such as myocardial infarction or stroke, a risk ratio is calculated to see if there is a difference in treatments. The benefit of a MA is that with an increasing number of studies, there is an increase in statistical power.

Care must be taken in deciding which studies to include in the MA. To increase the validity of subsequent MAs, drug developers can design prospective trials to be combined later. Attention must be paid to how the end-points are collected and that entails using the same definition for events, including technologies with the same or similar precision.

Conclusions

Proving cardiovascular safety in the drug development program can be difficult and lead to increased cost. NDAs can be delayed without a comprehensive and well thought out plan. New and evolving guidelines can be daunting to drug developers who would otherwise not concern themselves with cardiac safety. It is important to understand not only the drug under development and the class effects, but the regulatory landscape as well.

Timothy Callahan,* PhD, is Chief Scientific Officer, e-mail: tcallahan@biomedsys.com, and Pierre Maison-Blanche is Chief Medical Officer, both at Biomedical Systems, St. Louis, MO.

*To whom all correspondence should be addressed.