click to enlarge Figure 1. Simultaneous imaging of Vm and [Ca2+]i in a Langendorff-mode saline-perfused rat heart. (A) Vm (red) and [Ca2+]i (green) fluorescence signals (camera signals on a 16-bit scale) taken from a spot on the left ventricle (white-square). (B) Normalized fluorescence intensity maps (colorbar shown) at progressive time points during sinus rhythm. The delay of the [Ca2+]i relative to the Vm peak is clearly visible. Scale bar = 5 mm. (All figures: Aquoros LLC)

Most drugs never make it to market, failing during clinical trials. High attrition rates are frequently due to the harmful effects compounds can have on the heart’s electrical activity and function. Another common cause of failure is a lack of efficacy that is not identifiable until late in the pipeline, reflecting differences between in vitro models and clinical reality. Given the growing pressure to contain drug discovery costs, developing more efficacious preclinical development models is paramount.

A target for both safety profile testing and drug development is the heart. The safety and efficacy assessment methods of new drugs on the road to human trials includes in silico models, in vitro models, and complex animal models.

The major role of in vitro testing and computational modeling is to provide low-cost, high-throughput screening. However, time and cost savings mean little when it comes to investigating the clinical condition, as these models are limited in their clinical applicability.¹ There is a need to develop assays that provide the same high-throughput capacity of in vitro systems while using more relevant in vivo model systems that could even be applied to Phase 1 clinical trials. The closest example of a tool that meets these requirements is the electrocardiogram.

Current safety assessment methods
Cardiac safety profiling is ubiquitous across drug development, as safety profiling for cardiac arrhythmia potential is critical. For instance, an approved anti-histamine drug, terfenadine, was found to cause arrhythmias leading to sudden cardiac death.² The evaluation of COX-2 inhibitors further demonstrates research areas where animal model systems provided misleading results.³ In addition, despite initial in vitro success, developing heart drugs while relying on in vitro and primitive animal models has been fraught with failure. An important and prominent example comes from the ischemia reperfusion injury field. While research has led to several promising cases of potential blockbuster targets, all failed in advanced testing because of ineffectiveness.⁴

click to enlarge Figure 2. In vivo rat whole-heart preparation. (A) A schematic of the rat cardiopulmonary bypass circuit (CPB) based on a standard pump circuit (EJV: right external jugular vein catheter, VR: venous reservoir, RP: roller pump, MO: miniature membrane oxygenator, FA: right femoral artery catheter, AA: ascending aorta catheter via right carotid artery). In CPB mode, blood is pumped from the EJV through the MO to the FA, depicted by arrows in the figure. Dye is injected through the AA catheter. (B) Whole-animal view of the preparation. (C) Zoomed-in view of the open chest, with the heart in clear view (EL: esophageal ECG lead, ET: endotracheal tube). (D) An example of rapid ventricular pacing at 12 Hz to drop blood pressure (BP), after which dye was injected via the aortic root. The BP recovered soon after cessation of pacing. (E) Zoomed-in view of the heart immediately after rhod-2(AM) injection demonstrating a fluorescent coronary angiogram.

Proper cardiac excitability is the cornerstone of normal heart function. Research has shown that the cardiac action potential and evoked calcium transient are key pathophysiologic parameters of relevance to this process. This has proven true whether the goal is to investigate the arrhythmic potential of non-cardiac medicines, the mechanism of action for new anti-arrhythmic medications, or the role of calcium homeostasis in the development of novel therapeutics for heart failure. For this reason, the electrocardiogram has become a mainstay of cardiac research and clinical practice. However, it acts as a poor surrogate, providing only a summed version of the underlying electrical activity. A system capable of accurately and rapidly measuring all aspects of cardiac excitability at both the cellular and summed whole heart level would allow high-throughput safety and efficacy profiling within the heart with hitherto unseen accuracy. Such a system should be scalable and capable of providing similar data in vitro and in vivo.

Mapping the heart
Clinical arrhythmia diagnosis relies on electrophysiological testing, from simple intracardiac electrogram measurements to complex spatially-resolved electroanatomic mapping (EAM). EAM measures the electrical activity of tissue point-by-point while registering position in 3D space. This creates a tissue surface map of electrical activity. By comparison, basic science investigations of normal electrical activity and arrhythmias have relied on the higher spatiotemporal resolution method of optical mapping, which uses fluorescent dyes to image both the anatomy and electrophysiological parameters such as transmembrane voltage (Vm) and intracellular calcium ion concentration ([Ca2+]i).

While traditional optical mapping methods of these two key electrophysiological parameters have had a tremendous impact on arrhythmia research and drug action, such methods have remained relegated to the isolated (i.e. explanted) heart. Although useful, the traditional Langendorff-perfused isolated heart preparation is unphysiological in many respects. For instance, in this system, the heart is explanted from the animal, disconnecting critical neuro-humeral feedback and hormonal influences. It is then perfused with saline, a condition lethal to the live animal. The generated data from such heart preparation offers little clinical relevance.

The lack of progress in in vivo cardiac excitability mapping may in part be due to the experimental and technological challenges of traditional in vivo heart imaging methods. For example, even newer cardiac imaging modalities such as cardiac magnetic resonance imaging (cMRI), cardiac computed tomography (cCT), or ultrasound, do not capture relevant electrophysiological data. The first and only recorded use of optical mapping in vivo was reported in 1998.⁵ However, the research team was limited to studying a single parameter (voltage) using a low-resolution imaging system and complex illumination system based on lasers and acousto-optic deflectors. Thus, there is a critical need to develop an in vivo method that captures electrical excitability.

Changing the view

click to enlarge Figure 3. In vivo imaging of Vm and [Ca2+]i dynamics in rat ventricles/atrium, during sinus rhythm and in atrial fibrillation. (A) Normalized Vm fluorescence intensity maps (colorbar shown) at progressive time points of the cardiac cycle with the heart in sinus rhythm. (B) Normalized [Ca2+]i fluorescence intensity maps at sequential camera frames in part of the left and right atria during atrial fibrillation, induced by global ventilatory hypoxia. Scale bar = 5 mm.

The technological advances over the past 10 to 15 years are making it possible to develop sophisticated in vivo optical mapping tools. Such advances include the development of second-generation voltage-sensitive dyes with enhanced photostability and emission spectra in the near-infrared to optimally image in the presence of blood.⁶ These dyes can be thought of as novel contrast agents in development. While similar to more traditional iodine- or gadolinium-based contrast agents, these dyes can also report cardiac excitability.

High-speed electron multiplying charge coupled device (EMCCD) cameras have reached frame rates that permit simultaneous multi-color imaging using a single detector, making possible the development of scalable multi-parametric optical mapping systems.⁷ The authors used the Evolve 128 camera from Photometrics, Tucson, Ariz., for their research. With a 10-MHz readout, the Evolve 128 was designed for high-speed image visualization. When dealing with commonly desired cardiac physiological parameters, achieving very high speeds is not as much an issue as with other set-ups.  Thus, the authors could manipulate the speed by sharing the frames between samples, conducting the research with only a single camera.

Modern LEDs, ranging from the ultraviolet to the infrared, can now be used as excitation sources for all dyes used to study cardiac electrophysiology.⁸ Building on these recent advances, the authors are developing systems not only amenable to in vitro preparations, but also in vivo mammalian models. Such a system would be ultimately applicable to humans.⁹

Proof-of-concept system
A proof-of-principle study successfully demonstrated simultaneous action potential (AP) and calcium transient (CaT) imaging of a rat heart in vivo through 100% blood. A lab rat was used along with voltage and calcium dyes previously used in partial-blood perfused isolated Langendorff hearts. Figure 1 shows results obtained on an isolated saline-perfused rat heart to validate the single camera long-wavelength membrane voltage (Vm) and intracellular calcium ([Ca2+]i) imaging system. After validating the system on the isolated heart preparation, the imaging system was then applied to rat hearts in vivo.

click to enlarge Figure 4. In vivo imaging of Vm and [Ca2+]i dynamics in rat ventricles during sinus rhythm. (A) Vm (red) and [Ca2+]i (green) fluorescence signals taken from a spot on the left ventricle (white-square). (B) Normalized fluorescence intensity maps (colorbar shown) for Vm and [Ca2+]i during sinus rhythm; note the previously mentioned delay between Vm and [Ca2+]i peaks. Scale bar = 5 mm.

Figure 2 shows the in vivo preparation used. After anesthetizing the rat, a standard sternotomy, as is used for patients during bypass surgery, was performed. For some experiments, a cardiopulmonary bypass circuit was used to maintain coronary perfusion pressure. To load fluorescent dyes, it would have been ideal to use traditional coronary catheterization, but because of the small size of the rat, dyes were injected into the ascending aorta via the carotid artery and towards the coronary circulation. To accomplish this, blood pressure was reduced via rapid ventricular pacing for brief periods to decrease resistance to dye injection. Although heart movement reduced the signal quality, it was still possible to capture excitation waves traveling across the heart surface in vivo. However, movement will be much less of a problem in a large animal such as a pig or human where existing methodologies developed within cardiology to combat motion can be applied.

Figure 3 demonstrates in vivo imaging of Vm and [Ca2+]i dynamics in rat ventricles and atrium, during sinus rhythm and in atrial fibrillation. Figure 4 shows an example of the first reported measurements of Vm and [Ca2+]i in vivo.

From research to drug discovery to the clinic
This type of high-speed electrophysiological imaging using fluorescent probes has yielded tremendous insights into cardiac disease, and its potential in drug discovery seems highly promising. The authors are focusing their efforts in part on the translational goal of adapting their system to a minimally invasive and clinically compatible trans-catheter endoscopic system, which will enable high spatiotemporal electrophysiological studies of the heart in live animals.

In parallel, they are directly working with the pharmaceutical industry to develop a high-throughput drug testing imaging system that has the potential for bench-to-bedside data generation. Specifically, they hope their system will provide higher quality electrophysiology data in a more rapid fashion than currently available. This system can be applied across all model systems and potentially humans to identify systematic errors much earlier in the development pipeline clarifying the road to discovery. The hope is that this will allow a shorter, more efficient, and less expensive pathway for efficacious drug discovery.

About the author
Peter Lee is a doctoral student at Oxford University. Christopher Woods completed his clinical cardiology fellowship in the division of cardiovascular medicine at Stanford University and is currently a practicing clinical cardiologist. Woods and Lee recently co-founded Aquoros LLC, which develops in vitro and in vivo instrumentation for probing electro-mechanical properties of cardiac tissue.

References
1. Singh SS. Preclinical pharmacokinetics: an approach towards safer and efficacious drugs. Curr Drug Metab. 2006;7(2):165-82.
2. Pratt CM, et al. Risk of developing life-threatening ventricular arrhythmia associated with tefenadine in comparison with over-the-counter antihistamines, ibuprofen and clemastine. Am J Cardiol. 1994;73:346-352.
3. Pippin JJ. The Need for Revision of Pre-Market Testing: The Failure of Animal Tests of COX-2 Inhibitors. FDA Open Public Hearing: Arthritis Advisory Committee Drug Safety and Risk Management Advisory Committee (February 17, 2005).
4. van der Worp HB, et al. Can Animal Models of Disease Reliably Inform Human Studies? PLoS Med. 2010;7(3):e1000245.
5. Dillon SM, et al. A system for in-vivo cardiac optical mapping. IEEE Eng Med Biol Mag. 1998;17:95-108.
6. Matiukas A, et al. Near-infrared voltage-sensitive fluorescent dyes optimized for optical mapping in blood-perfused myocardium. Heart Rhythm. 2007;4:1441-1451.
7. Lee P, et al. Single-sensor system for spatially resolved, continuous, and multiparametric optical mapping of cardiac tissue. Heart Rhythm. 2011;8:1482-1491.
8. Entcheva E, et al. Fluorescence imaging of electrical activity in cardiac cells using an all-solid-state system. IEEE Trans Biomed Eng. 2004;51:333-341.
9. Lee P, et al. in vivo optical mapping of voltage and calcium in the heart. Submitted.