The Heart of the Matter
Safety Assessment
Regina Kelder

The Heart of the Matter

Regulators want to swap the QT test for a 21st century approach for assessing the cardiac safety of novel drugs. Is it likely that the platform will meet the envisaged 2015 deadline?

Nearly a century ago, the medical profession identified a potentially lethal type of drug-induced heart arrhythmia. Back in 1922, the analysis of 460 patients receiving quinidine to control an erratic heartbeat found that this drug unexpectedly caused eight patients to lose consciousness. This phenomenon remained extremely unusual, at least in the medical literature, until 1958 when the schizophrenia drug thioridazine was found to cause the same severe adverse effect. Thereafter, the examples mounted.

By the mid-1990s the number of noncardiac drugs associated with similar cardiac effects (most likely torsades de pointes arrhythmia (TdP)) had expanded so dramatically that a European regulatory panel recommended all drug candidates be screened routinely for this cardiotoxic activity. By 2005 global standards were in place for identifying drug-induced TdP effects. These International Conference of Harmonization (ICH) S7B and E14 safety guidelines designed to determine whether a drug delays ventricular repolarization, were subsequently applied by drug companies in Europe, Japan, the U.S., and elsewhere.

Specifically, the biological tests investigate whether and how much a drug dose extends the interval between the onset of the Q wave and the end of the T wave in the cardiac electrical cycle. The longer that the time interval between the Q and T waves is prolonged by a drug treatment, the more likely it is that the drugs bind to and inhibit the specific function of proteins (ion channels) that govern cardiac ion channel cycling. Indirectly, certain drugs have also been known to affect this function by dysregulating the ion channel synthesis.

Numerous clinical studies show that in vivo QT assays do a reasonably good job identifying drugs that delay ventricular repolarization. However, the assays aren’t foolproof. Importantly, they cannot generate reliable metrics for the actual proarrhythmic risk of a drug in patients, which is one reason why about a quarter of drugs that made it to market between 1990 and early 2012 were withdrawn because of some unintended drug-induced cardiac proarrhythmic risk that the QT assays is not designed to adequately capture, if at all.

Which brings us to the focus of this blog—the Comprehensive In Vitro Proarrhythmia Assay (CiPA), a novel safety screening proposal that consists of two in vitro biological assays and onein silico computational analysis. Dr. Norman Stockbridge, who directs the US Food and Drug Administration’s Office of Drug Evaluation  I-Division of Cardiovascular and Renal Products (DCaRP), would like to see the CiPA platform replace the ICH’s 2005 S7B guideline by July 2015. But, as Icilio Cavero and Henry Holzgrefe, experts in cardiac electrophysiological safety of drugs pointed out in their commentary last month in the journal Expert Opinion on Drug Safety, there are a number of factors to consider before implementing what they, nonetheless, describe as a visionary initiative for the 21st century. Regina McEnery, the senior scientific writer who oversees the Eureka blog, chatted recently with Holzgrefe, a safety pharmacology consultant for Charles River and the author of more than 75 scientific abstracts and peer-reviewed publications focused on cardiovascular safety assessment of small molecules and biologics, about the scope and timing of CiPA.

What are the main limitations of the S7B and E14 guidelines?
Drug developers need to demonstrate either the absence of effect or an evident cardiac proarrhythmia liability. A ball park estimate to conduct a thorough QT study is about US $2 million. Along with the preclinical studies, you then have to also enroll a sufficient number of patients in clinical trials that may involve multiple sites, have the data analyzed, and then submit it to the FDA. With the current paradigm, drug developers incur millions of dollars in additional direct costs while delaying the timely marketing approval of potentially beneficial drugs, assuming that the tests exclude a potential cardiac risk. The FDA and Dr. Stockbridge realized that the current approval strategies squandered time and resources, so they proposed a drug evaluation safety strategy, named CiPA, that would be more cost-effective and less reliant on intact animal testing.

The CiPA method involves also stem cell technology, correct?
Yes. The induction of pluripotent stem cell-derived cardiomyocytes (CMs) from individuals has opened the door to modeling the cardiac electrical currents. The latter underlie the risk of cardiac arrhythmias in assays performed in CMs from the healthy, adult human heart. Huge quantities of human ventricular CM proteins can now be obtained by heterologous expression using standard mammalian cell lines. What Dr. Stockbridge and the FDA propose in the CiPA initiative is to use ion channels in vitro that are known to mediate QT prolongation and other cardiac proarrhythmic indices, and then characterize potential drug effects on the ion channels contributing to the shape of the cardiac electrical activity, called action potential.

Which ion channels are being targeted in the CiPA proposed assays?
The first core component of CiPA addresses the effects of candidate drugs on the main sodium (Na+), calcium (Ca2+) and potassium (K+) cardiac ion channels, which govern the action potential.  These are the proteins responsible for the depolarization and repolarization, and when they are functioning properly the likelihood for drug-induced proarrhythmic liability is virtually nil.

What other components round out the CiPA strategy?
There is also in silico modeling involving complex computational models that can be used to characterize or predict toxicological outcomes from the reconstruction of the action potential which incorporates the effects of the ion channel dataset. CiPA proposes to use this technology to predict drug-induced changes in the cardiac action potential (AP) simulations. The in silico findings would then be verified by determining the associated electrical activity in human stem cell-derived ventricular CMs, where in vitro assays would be used to confirm the effects of a drug candidate revealed by the in silico simulation.

What are your main concerns about CiPA?
The CiPA initiative certainly represents the cutting edge of safety assessment. My major concern is about meeting the proposed July 2015 timetable. Multiple validation studies need to be completed to ensure that CiPA is an adequate replacement for the traditional QT assay. This is going to have a massive impact on laboratories [like Charles River] that conduct these kinds of safety screenings. There will have to be an international infrastructure put together to discuss and agree on how CiPA should be implemented and how it should be validated. CiPA is intended to replace an ICH standard that the European, Japanese, and US regulatory authorities now scrupulously apply. In order to be accepted on an international level, CiPA needs to be performed exactly the same in Tokyo as it is in Paris or Reno. It has to be validated internationally with the same reference compounds. Absent these assurances, the ICH guidance will remain in place as the international community will not accept CiPA as a replacement strategy with anything less than full cross-site validation. Yet here we are, in July 2014, and very little of this requisite work has been completed.

The expert opinion article offers some suggestions on how to optimize the CiPA assay evaluating a drug’s effect on key cardiac ion channels. What are you proposing and why?
While three main ion channels (Na+ ,Ca2+, K+) contribute to the cardiac action potential, there are others that play an essential role in the drug cardiac safety assessment. If any of these additional cardiac currents are suspected off-target sites of action of a drug candidate, they should also be a matter of CiPA scrutiny. A way to optimize the assay is to implement reliable, fast, inexpensive automated patch-clamp technologies to study ion channel effects. However, at present, the available platforms need improvement. These are not just our observations. The Expert Committee responsible for proposing ion channel investigation protocols for the CiPA initiative has also made this observation.

You also had some concerns about the stem cell CM technology.
Yes, sources of stem cell-derived CMs have not been adequately well-characterized in regulatory assays. There is a lot of drift across laboratories. Indeed, current commercially available human stem-cell derived CM cell lines are not highly pure populations of ventricular CMs and they don’t appear to fully recapitulate the electrophysiological features of native healthy human CMs and have not always responded, as expected, to established proarrhythmic agents, raising questions about the human predictability of the assays. CiPA is an excellent initiative which will require the effort and the enthusiasm of all stakeholders engaged in ensuring the nonclinical cardiac safety of drugs to become a reality.