The ABCs of Thorough QTs

In the last decade, cardiac safety concerns have been responsible for more drug withdrawals than any other safety reason. In response, both regulators and drug companies now have a renewed focus on TQT trials, which help assess the cardiac safety of drugs.

The ICH's E14 Guidance increased the demands on drug developers and made the regulatory process even more complex. Major regulatory agencies around the world, including the FDA, EMA, MHLW (Japan), and Health Canada, require a TQT study not only for new drugs submitted for approval, but also for any change to an approved drug that significantly increases its exposure. In fact, some drugs that were originally approved in the 1960s have recently undergone TQT trials because of changes in dosage.

BEN EDWARDS/GETTY IMAGES

To conform with the E14 Guidance, a TQT trial must meet the following criteria:

• Conducted in humans

• Robust enough to detect a 5 millisecond prolongation of the QT interval

• Placebo-controlled

• Use a positive control to validate the study

If the TQT study is negative, the drug can proceed normally in development. However, if the TQT study is positive, all subsequent studies will need extra ECG monitoring and the drug may be required to include specific instructions on its label, ranging from special dosing instructions to black box warnings.

Critical elements of TQT design

TQT studies are an essential element of any drug development program. However, because of their importance to the progression of the drug development program and the cost of conducting the study, TQT studies must be conducted with careful design considerations.

Some of the important aspects of the design include the number of subjects, timing of the ECGs, and the amount of the drug dosed. The number of subjects required for a TQT study can be simply determined by the variability of the QT measurement and the power needed to achieve the required QT prolongation. However, there are a number of considerations that go into determining the variability.

The first variability factor is the study design: crossover versus parallel. A crossover design benefits from the reduced variability and increased power of the subjects acting as their own control. Parallel design studies are better suited for drugs with long half-lives, when dosing within a patient population, and for drugs with known period effects.

A second factor affecting the variability of the study is the number of groups required. In order to complete TQT studies as quickly as possible, many companies split the study into smaller groups conducted at different sites within the same contract research organization (CRO) or among different CROs. Different sites increase the variability in the QT measurement.

The final factor affecting the variability is the timing of the baseline ECGs. Baseline ECGs should be taken at multiple times throughout the day before dosing; these times of day should correspond to the times of day the ECGs are measured following dosing. In some cases, baseline ECGs are measured for more than one day before dosing.

In terms of ECG measurement after dosing, the collection of ECGs during the TQT study should be such that the Cmax of the drug is carefully captured, unless there is reason to believe the maximal QT prolongation of the drug will not occur during the Cmax. Taking pharmacokinetic (PK) samples to correspond with the ECG is also recommended, as the exact drug concentration seen during the ECG time points can be determined. Furthermore, the PK samples will facilitate pharmacokinetic/pharmacodynamic (PK/PD) modeling, which may help researchers understand whether or not a drug causes QT prolongation. This understanding may prevent the need to conduct additional studies.

The dose levels used in a thorough study are very important. The E14 guidance makes it clear that in most cases, the TQT study should not be dosed solely at the therapeutic dose, but also at a higher dose where the exposure of the drug may be similar to that seen after accumulation of the drug or because of drug-drug interactions. Therefore, a TQT study should be conducted as soon as the therapeutic dose and pharmacokinetics of the compound are well known. Ideally, the TQT study would occur as early as possible in the development process, since the result of the TQT study will influence the design of Phase III studies.

Finally, in most cases, the TQT study is conducted at steady-state, again to maximize the exposure. For drugs with short half-lives, this may not be required.

TQT timelines

The design of TQT studies should begin approximately six months in advance.

• More than six months before dosing. Check investigational new drug application (IND) status. If the initial studies on the drug were conducted in the United States, an IND will already be open. If there is no IND and the TQT studies are to be conducted in the United States, an IND must be filed. Start the evaluation process for selecting a CRO. Site visits are highly recommended, as the physical setup of the clinical pharmacology unit (CPU) is extremely important in TQT studies.

• Six months before dosing. Define the therapeutic and supratherapeutic doses. The therapeutic dose will be defined in a Phase IIa study conducted in patients or in healthy subjects if an appropriate biomarker exists to measure the drugs' pharmacodynamic action. The supratherapeutic dose is "substantial multiples" of the therapeutic dose and is used to represent the "worst-case scenario" that might be seen with accumulation and inhibition. At this point a final decision regarding a CRO can be made.

• Three months before dosing. Submit the study design to the FDA for feedback. The FDA has established a fast review process to ensure timely responses. Some CROs can assist with the submission to the FDA and suggest the types of questions to ask the FDA.

• Two months before dosing. Incorporate feedback from the FDA into the study design. Review options for selecting patient populations, if required.

• One and a half months before dosing. Submit the final protocol to the institutional review board (IRB) for approval.

• The day of dosing. Have a staff member from the sponsor organization observe dosing.

• One and a half months following first dose. Complete the clinical component of the TQT. This time frame is variable, as it depends on the number of days of dosing and the washout needed between periods. The ECG readings will begin at the core ECG laboratory. Begin bioanalysis of the PK samples: The samples must be measured even if the PK of the drug is well-known so that any potential effect on the QT interval can be compared to the actual Cmax of the subjects. Also, having the PK data will enable PK/PD modeling.

• Three months after dosing. Receive draft final report for the TQT study from CRO. Following approval, the report is ready to be finalized and submitted to the FDA. As TQT studies are strongly suggested to have positive control arms, it is vital that a 5 millisecond significant difference is seen in this arm.

Common challenges in TQT studies

There are several challenges that can present obstacles when designing TQT trials. It is very important to address and consider these issues in the design stage of a TQT study, and when selecting the CRO to perform it.

Inappropriate doses. The E14 Guidance indicates that a positive control arm is desired and asks that one arm of the study be at the therapeutic dose and one arm at a significantly higher, or supratherapeutic, dose. An appropriate range for the supratherapeutic dose is often a dose within 10 times of the therapeutic dose. However, the safety profile of many drugs does not allow a dose of this size. Dosing a study with a supratherapeutic dose that is too high can result in significant adverse events that do not permit subjects to complete the study and would result in the loss of complete data for the dose level.

Conducting a maximum tolerated dose (MTD) study allows a sponsor to clearly define the dose level that will be tolerated. MTD studies are conducted in a small group of subjects and can be completed quickly. Therefore, the cost of the MTD study can easily be justified when compared to the cost of repeating a TQT study.

The guidance also states that, "if the concentrations of a drug can be increased by drug-drug or drug-food interactions involving metabolizing enzymes (e.g., CYP3A4, CYP2D6) or transporters (e.g., P-glycoprotein), these effects could be studied under conditions of maximum inhibition." In this case, an understanding of the metabolic pathway and the resulting drug-drug interactions (DDI) are required. If the preclinical data indicate metabolism at one of the key CYP enzymes or transporters, a DDI study will allow determination of how this interaction affects the concentrations of the drug. A dose that results in a similar exposure as that seen with the maximum clinically indicated dose under the interaction should then be used as the supratherapeutic dose.

QT/RR hysteresis. Another pitfall is not taking into account QT/RR hysteresis, which refers to the delay between changes in heart rate and QT interval when designing or conducting studies. QT/RR hysteresis is not corrected by the standard QT correction factors.

A change in heart rate—due to external factors such as stress or even drug-inducement—can take up to four minutes to result in a change in the QT interval. This lag is not consistent across subjects, ranging from a few seconds to several minutes.

The TQT trial's clinical pharmacology facility needs to be equipped with staff trained to maintain the most-consistent heart rate possible, so the likelihood of QT/RR hysteresis is minimized and the usual correction factors (e.g., Bazett and Fridericia) can be relied on.

Experienced CROs will monitor several factors to minimize QT/RR hysteresis, including:

• Timing of blood draws so that they occur after ECG measurement

• Limiting the activity of subjects

• Reducing the excitement, and therefore subjects' heart rate, from items such as horror movies

• Restricting use of computers, cell phones, and other devices

• Clearly explaining what is expected of the subjects well in advance of the trial

It is also important to consider the effect the drug or administration of the drug may have on the heart rate. A large number of drugs are known to increase or decrease the heart rate. However, even drugs with no known effect on heart rate can result in a heart rate change if they are intravenously infused with a significant volume of vehicle. By controlling the changes in heart rate, the variability seen in the QT intervals will drop and will result in the need for fewer subjects to maintain sufficient power for the study.

Insufficient power to detect change in the positive control arm. Another challenge is in the calculation of the power of the TQT prolongation study. The positive control arm must have a significant QT prolongation change. It is very important to ensure the number of subjects is sufficient, taking into account the known change in QT of the chosen positive control and the variability expected in the study.

Factors that can impact variability include proper baseline correction, ECG machines used, and other design elements. Without a good estimate of the variability of the study, it is not possible to provide a good estimate of the sample size. It is critical to work with an experienced clinic and central ECG provider that has a solid understanding of all of the components of the overall variability.

How to choose a CRO for a TQT study

As in all clinical research, proper study design and first-rate administration of a TQT study are essential. Proper scientific and practical experience is especially important. Even larger pharmaceutical companies can benefit from the design expertise of an experienced CRO. However, for smaller companies that may have limited experience with TQT studies, collaborating with an experienced CRO is essential. In terms of designing a TQT study, theoretical knowledge is not sufficient. It is important to look for a CRO with expertise in the actual performance of TQT studies, as this experience leads to the best possible design for the conduct of a trial.

Once the design of the TQT study has been determined, the focus changes to conducting the study itself.

Here, two aspects are important: the staff and the physical setup of the CPU within the CRO. These two factors help determine the variability in the QT measurements that will be seen in the study. As the variability goes up, the chance of having a significant signal in the positive control arm goes down, unless the subject numbers are increased to compensate. How do these factors affect variability? It is well known that the correction factors for heart rate are far from perfect. Therefore, the best way to reduce the variability introduced by changing heart rate is to not rely on correction factors, but to do everything possible to keep a subject's heart rate consistent throughout each ECG measurement. A good way to maximize staff expertise in conducting TQT studies is to dedicate individuals to these types of studies, ensuring that the experience is concentrated among a select group of key personnel.

The physical setup of the CPU also plays a role in the changes seen in the subject's heart rate and, therefore, the variability of the study. A well-designed CPU is set up so that all the subjects can be closely monitored during the study. Furthermore, the subjects should be able to remain comfortably supine throughout the ECG measurements. Comfortable hospital beds, cleanliness, and lighting are all important.

The most important aspect of the CPU is the group size that it can handle. The typical TQT study requires a large number of subjects. If the CPU can only handle small groups, then many different groups will be needed. Dosing of multiple groups can delay the program and increase the cost.

Not every sponsor has the capacity to internally analyze the data from the TQT study and will need to outsource this work. If this is the case, it is often easiest to use a single CRO that can perform the clinical aspect of the study in addition to providing the required data management, pharmacokinetics, statistics and medical writing.

The selected CRO should have previously used combined services for a number of studies and have the ability to conduct in-house analysis and PK/PD modeling of the QT data using sophisticated population PK/PD approaches.

Experienced CPUs will have a good background working with a number of different companies and offering centralized ECG services to provide the flexibility to adapt to the different equipment and software. Finally, in looking for experience, note not only the number of studies a CRO has performed, but also how recently and how frequently TQT trials were conducted.

TQT trials are extremely complex. However, they are also necessary for drug approval in most countries around the world. Understanding the critical elements, timeline, and common pitfalls of these trials and selecting the best CRO partner will help ensure sponsors achieve accurate results in the most efficient manner.

Alan Copa, PhD, is President of Cetero Research, Clinical Operations Services, 4801 Amber Valley Parkway, Fargo, ND 58104, e-mail: Alan.copa@cetero.com