|Home | About | Journals | Submit | Contact Us | Français|
To explore the feasibility of targeted imaging of the angiotensin II subtype 1 receptor (AT1R) in cardiac tissue, using clinical hybrid positron emission tomography/computed tomography (PET/CT).
AT1R is an attractive imaging target due to its key role in various cardiac pathologies, including post-infarct left ventricular remodeling.
Using the novel AT1R ligand [11C]-KR31173, dynamic PET/CT was performed in young farm pigs under healthy conditions (n=4), and 3–4 weeks after experimental myocardial infarction (n=5). Ex vivo validation was carried out by immunohistochemistry and PCR. First-in-man application was performed in 4 healthy volunteers, at baseline and under AT1R blocking.
In healthy pigs, myocardial KR31173 retention was detectable, regionally homogeneous, and specific for AT1R, as confirmed by blocking experiments. Metabolism in plasma was low (85±2% of intact tracer after 60min). After myocardial infarction, KR31173 retention, corrected for regional perfusion, revealed AT1R upregulation in the infarct area relative to remote myocardium, while retention was elevated in both regions when compared to myocardium of healthy controls (8.7±0.8 and 7.1±0.3 vs 5.8±0.4 %/min for infarct and remote vs healthy controls; p<0.01 each). Postmortem analysis confirmed AT1R upregulation in remote and infarct tissue. First-in-man application was safe, and showed detectable and specific myocardial KR31173 retention, at an albeit lower level than pigs (LV average retention: 1.2±0.1 vs 4.4±1.2%/min for humans vs pigs; p=0.04).
Noninvasive imaging of cardiac AT1R expression is feasible using clinical PET/CT technology. Results provide a rationale for broader clinical testing of AT1R-targeted molecular imaging.
Scientific work in recent years has highlighted the role of the renin-angiotensin system (RAS) in cardiac pathology1–3. In addition to the circulating RAS, which contributes to the systemic regulation of global cardiovascular homeostasis, the heart has an intrinsic RAS which mediates loco-regional mechanisms such as interstitial fibrosis, myocyte hypertrophy and apoptosis4. Activation of the intrinsic myocardial RAS may contribute to changes of geometry, structure and function which are hallmarks of heart failure progression.
Interest in targeted imaging of maladaptive mechanisms contributing to heart failure and left ventricular remodeling is increasing5. Novel molecular imaging techniques may not only improve pathophysiologic understanding. They may also provide prognostic value and refine therapy. In this context, the myocardial RAS appears to be an attractive target. Using ligands for the primary RAS mediator in myocardium, the angiotensin II type 1 receptor (AT1R), studies showed that regional AT1R upregulation can be visualized noninvasively in rodents after myocardial infarction6, 7. But inter-species differences of RAS have been reported8, 9, and the usefulness of cardiac AT1R imaging in large mammals and humans remains to be demonstrated.
Accordingly, we sought to explore the translational potential of myocardial AT1R imaging, by using clinical hybrid positron emission tomography/computed tomography (PET/CT).
The study protocol was approved by the Johns Hopkins Institutional Animal Care and Use Committee. Animals were maintained according to principles of the American Physiological Society. Nine female young farm pigs (20–30kg) were enrolled. For experimental procedures, animals were held under general anesthesia (induction with ketamine hydrochloride 200–400mg, maintenance with 1.2–2.0% isoflurane). For myocardial infarction, coronary catheterization was performed as previously described10 in 5 animals. Balloon occlusion of the mid left anterior descending artery was performed for 120minutes. Post-operative treatment included narcotics and non-steroidal anti-inflammatory drugs. Hybrid PET/CT was performed 3–4 weeks later. Four animals underwent PET/CT under healthy conditions.
PET/CT was conducted using a GE Discovery Rx VCT scanner (GE Healthcare, Waukesha, WI). Pigs were anesthetized and positioned supine in a cradle. A CT scout scan was followed by low-dose CT for attenuation correction. [11C]-KR31173 was administered intravenously via ear vein (300–500MBq). List-mode acquisition (60min) was started simultaneously. Venous blood was taken at 20, 40 and 60min, to determine plasma metabolites by column-switch high performance liquid chromatography11. After waiting for radioactivity decay (100min after injection), another low-dose CT was obtained, followed by intravenous [13N]-ammonia (NH3) infusion (370–555MBq) and 20min list-mode acquisition. Next, a delayed contrast-enhancement CT was performed as described14, using 70mL Visipaque (GE Healthcare) and helical, retrospectively gated acquisition 10min after injection.
In 3 healthy pigs, PET/CT was repeated under blocking conditions (SK-1080, 2mg/kg IV, 30min before [11C]-KR31173 injection)12.
List-mode data were resampled to attenuation corrected, iteratively reconstructed, tomographic images. Alignment of CT and PET was checked using fusion software15. Static images were obtained for KR31173 (30–60min) and NH3 (10–20min). ECG-gated images were obtained for NH3 (8 bins), and dynamic images were obtained for both agents (27 frames for KR31173: 12×10, 6×60, 4×180, 2×300, 3×600s; 21 frames for NH3: 12×10, 6×30, 3×300s).
Static PET data were volumetrically sampled, and myocardial polar maps were generated. A threshold of 60% of the individual maximum was used to define perfusion defect16.
For quantitative analysis, myocardial segments defined from static NH3 images were applied to dynamic series, and time-activity curves were obtained. Additionally, the arterial input function was defined by a small region of interest in left ventricular cavity. Absolute myocardial blood flow at rest was quantified using a 3-compartment model for NH317. For KR31173, a myocardial retention index (KR-ret) was determined according to the following formula:
From gated NH3 datasets, resting left ventricular ejection fraction (LVEF) and end diastolic volume (EDV) were measured as previously reported18.
Venous blood samples were obtained in animals before infarction, and again on the day of the PET/CT scan. Plasma levels of renin, angiotensin II and aldosterone were measured using commercial services (Quest Diagnostics®, Madison, NJ).
After imaging, anesthetized pigs were euthanized (4mmol/L potassium chloride IV). Hearts were removed and gross left ventricular short-axis slices were created10. Under guidance by TTC stain (triphenyltetrazolium-chloride, 12.5mL/kg, 2%), samples were collected from infarct, remote region and right ventricle. Samples were frozen at −80°C for polymerase chain reaction (PCR), or fixed in 10% formaldehyde for histology/immunohistochemistry. Thin slices were stained by Hematoxylin/Eosin (HE) or Masson-trichrome. AT1R immunohistochemistry was performed using goat polyclonal AT1R antibody (SantaCruz, Delaware, CA).
For quantitative PCR, total ribonucleic acid (RNA) was extracted from homogenized tissue (RNAeasy, Qiagen, Valancia, CA). Reverse-transcription PCR (1μg RNA) was performed using qScript™ cDNA Kit (Quanta BioSciences). Real-time quantitative PCR was performed with StepOnePlus™ (Applied Biosystems, Foster City, California, USA), using PerfeCTa® SYBR® Green FastMix®, ROX (Quanta BioSciences). GAPDH (glyceraldehyde-3-phosphate dehydrogenase) served as internal control. The sequences of the primers for porcine AT1R and GAPDH are specified in Table 119. According to the delta Ct method, expression of AT1R mRNA was normalized to GAPDH mRNA and related to sample measurements from non-infarcted myocardium. Specific amplification was confirmed by melting curve analysis. Experiments were performed in triplicate and repeated 3 times.
Approval for human application of KR31173 was obtained by investigational new drug application from the Food and Drug Administration, and from the Johns Hopkins Institutional Review Board. Safety was the primary focus. Accordingly, 4 healthy volunteers (male, age 24±2 years) were included. Cardiovascular or other disease was ruled out by clinical examination, history, blood testing and ECG. After written informed consent, subjects underwent KR31173 PET/CT. Injected activity was 650±54 (579–709) MBq, specific activity 139.3±66.9 (75–230.4) GBq/μmol, and injected cold mass 2.96±1.18 (1.53–4.22) μg of KR31173. Based on animal data, which suggested an effective dose of 9.05μSv/MBq, the individual effective dose was calculated to be 5.9±0.5mSv. Blood pressure, heart rate, and subjective symptoms were assessed before tracer injection, during imaging, and at the end of the scan. No NH3 or CT contrast agent were given. Otherwise, PET/CT acquisition and data analysis were performed as described above for animal studies. In two subjects, imaging was repeated on the subsequent day under AT1R blocking conditions, 3 hours after an oral dose of 40mg olmesartan.
All results are expressed as mean±standard deviation. Results within the same group were compared by paired t-test. Results between different groups were compared by unpaired t-test. Due to possible heterogeneity of variance, KR31173 retention in humans vs animals was compared by Welch t-test. All testing was performed using SPSS 13 (SPSS®) and PRISM ver.5 (Graph Pad®). P<0.05 was considered statistically significant.
Myocardial KR31173 was clearly visualized and regionally homogeneous, similar to NH3 uptake (Figure 1A). Strong liver uptake of KR31173 was present, but did not interfere with myocardial visualization. Kinetic analysis showed stable myocardial retention over time (Figure 1B). Global KR-ret was 5.5±1.58, 4.10±0.76, 3.62±0.51 at 10, 30, and 60min. Plasma metabolite analysis revealed tracer stability (89±1% of tracer intact at 10min after injection, 88±3% at 30min and 85±2% at 60min). Blocking resulted in complete disappearance of myocardial KR31173 uptake (Figure 1C). Quantitatively, KR-ret was reduced by >90% (0.39±0.10 at 30min; P<0.001).
Quantitatively, perfusion defect size comprised 26±6% of the left ventricle. CT delayed enhancement matched the perfusion defect (Figure 2), showing mostly transmural tissue damage associated with wall thinning, and non-transmural enhancement in a small borderzone. LVEF was 32±5%. Compared to healthy normal myocardium (0.40±0.14mL/min/g), global myocardial blood flow was significantly reduced in the infarct region (0.18±0.04mL/min/g; p<0.001), but not in the remote region (0.39±0.08mL/min/g; p=0.89) of infarct animals. Probably due to the general cardio-suppressive effect of anesthesia20, overall blood flow was low.
Systemic renin, angiotensin II and aldosterone were unchanged between pre- and post-infarct states. Plasma concentrations were 0.18±0.2 vs 0.15±0.12u/L, 15.0±2.0 vs 13.0±3.6ng/mL, and 2.9±1.4 vs 2.5±1.9ng/mL for renin, angiotensin II, and aldosterone, respectively (all p=ns).
Visual comparison of KR31173 and NH3 images suggests that KR31173 uptake in the hypoperfused infarct region exceeds residual regional perfusion (Figure 2). Notably, right ventricular uptake appears enhanced, too.
Comparison of absolute KR-ret with healthy controls suggests a global upregulation in infarct animals (Figure 3A). Uncorrected KR-ret at 30min was significantly elevated in the remote area of infarct animals (6.10±0.64%/min) when compared to infarct region (2.78±0.82%/min; p=0.002) and normal myocardium of healthy controls (4.10±0.25%/min; p<0.001). When KR-ret was normalized to regional perfusion in order to correct for partial volume effects and reduced flow in the infarct region (Figure 3B), the elevation of KR-ret in remote regions persisted, but regional KR-ret was most strongly upregulated in the infarct region (0.087±0.008 vs 0.071±0.003 vs 0.058±0.004 for infarct, remote and control; p=0.001 for control vs infarct, p=0.015 for infarct vs remote, and p=0.005 for remote vs control; Figure 3C).
Comparison of tissue samples (Figure 4A) confirmed in vivo findings. Quantitative PCR revealed significantly increased AT1R expression in infarct and remote regions compared to healthy controls (P<0.001 and P=0.016). Also, there was a trend towards higher expression in infarct versus remote region (P=0.057). Notably, AT1R expression in the right ventricle of infarct animals was highest (p<0.001 vs all others) (Figure 4B).
Histologic analysis showed necrosis and fibrosis on HE and Masson-trichrome staining in the infarct area, while both were absent in remote myocardium. Immunohistochemistry showed anti-AT1R antibody binding to spindle-shaped cells, presumably myofibroblasts, in the infarct region, while there was also diffuse binding to cardiomyocytes in remote areas (Figure 4C).
None of the human subjects reported and symptoms after KR31173 administration. Blood pressure (mean arterial pressure: 77±9mmHg before vs 77±9 during imaging) and heart rate (68±9 vs 65±7/min) remained stable. No adverse events were recorded.
At baseline, myocardial KR31173 retention was visually detectable and regionally homogeneous (figure 5). Similar to pigs, retention was stable over time and strong liver uptake was present. Absolute LV retention was 1.2±0.1%/min at 40min, which was significantly lower compared to healthy pigs (4.4±1.2%/min; p=0.04).
After blocking with olmesartan, myocardial retention disappeared and blood pool activity increased. KR31173 retention dropped from 1.3±0.1 to 0.7±0.1%/min. Residual retention after blockade was mostly attributable to myocardial spillover from elevated blood pool.
To our knowledge, this is the first report of noninvasive in vivo visualization of myocardial AT1R expression in large mammals and humans. Proof of safety and feasibility provides a rationale for further clinical testing.
On the myocardial tissue level, RAS activation stimulates cell growth, inflammation, fibrosis and apoptosis through AT1R21. Modest activation is thought to be part of the adaptive response to tissue damage and contributes to repair4. But an excessive, maladaptive response may contribute to tissue remodeling and loss of function. A diagnostic method which identifies AT1R upregulation on the myocardial tissue level may be of clinical value in several ways: First, if quantified early after acute myocardial infarction, the severity of upregulation may predict risk for subsequent heart failure development. Second, myocardial AT1R imaging may be useful in other situations such as left ventricular hypertrophy and hypertension, and it may further elucidate the link between renal and cardiovascular disease. And third, molecular imaging may be used for individual optimization of AT1R blockade, based on the level of receptor occupancy in target tissue. The present study provides a foundation for future testing of these hypotheses.
Of note, species differences of RAS/AT1R need to be considered. Prior work in rats6 and mice7 showed strong upregulation of AT1R in the infarct area, without imaging signal from remote or healthy myocardium. This is in contrast to pigs, where remote myocardium showed significant AT1R binding, and where regional upregulation in the infarct region was less pronounced. Prior experimental work confirms these differences between species8, 9, 22. Notably, humans, like pigs, showed detectable levels of myocardial AT1R at baseline, although absolute retention was significantly lower. Whether this is due to further species differences, or due to an effect of anesthesia in animals cannot be clarified.
The presence of upregulation of myocardial AT1R in the absence of changes of systemic RAS components emphasizes the value of noninvasive imaging of myocardial tissue. The existence of an intrinsic RAS in myocardial tissue is now well established4, 9, 23. Work by other imaging groups has recently focused on imaging of angiotensin converting enzyme (ACE) as another component of myocardial tissue RAS, which is upstream of AT1R24. This may add an additional perspective to RAS imaging, because multiple tracers may be combined for dissecting molecular mechanisms. In addition to developing tracers for other molecular components of RAS, probes for AT1R may also be optimized, e.g. by focusing on labeling with the broader available isotope fluorine-18, or by focusing on a reduction of high nonspecific liver uptake.
It is noteworthy that our study also showed significant elevation of AT1R in the right ventricle. This is along the line of prior work, where right ventricular angiotensin binding was also found to be enhanced after infarction9, but the finding was not discussed in detail. Differential alterations of loading conditions during anesthesia may be a contributor20, but the underlying mechanisms for this observation cannot be clarified from the present study. Although PET imaging of the right ventricle is not reliable due to partial volume effects resulting from thin wall, the role of the low-pressure circulation in RAS-related cardiac pathology deserves some attention in future studies.
Limitations of our work should be recognized: First, sacrifice of animals after imaging was necessary to obtain ex vivo validation. Our study thus lacks information about the time course of AT1R after myocardial infarction. Prior work in rats suggests a peak after 3 weeks6, which supports the timing of imaging in the present study. Second, ex vivo tissue workup shows a general limitation of in vivo imaging at relatively low spatial resolution: The imaging signal is not specific for cell type. As suggested by immunostaining in this and other studies7, the AT1R signal in the infarct region originates mostly from myofibroblasts25, while the signal from remote and healthy myocardium originates from myocytes. It is not clear, if and how the cell-specific origin of AT1R upregulation is relevant for healing and remodeling. But integrated imaging may help to overcome this limitation. Combination with CT delayed enhancement localizes the AT1R signal to scar or intact tissue. Of note, specificity of the tracer for AT1R has been shown previously not only in healthy myocardium, but also in infarcted myocardium or extracardiac tissue6, 13. Third, the employed pig model is not entirely representative for clinical post-infarct remodeling. General depressive effects of anesthesia may explain low LVEF and low blood flow20, the observation period is short, and there is absence of coronary atherosclerotic disease. Nevertheless, this pig model is established to study the biologic effects of tissue damage20, and our results provide a rationale for further clinical testing. Fourth, we employed a retention index for [11C]-KR31173 to quantify AT1R expression. While this approach corrects for the amount of available tracer, it is dependent of partial volume and does not provide truly quantitative receptor density. More sophisticated approaches are established for other tracer26, but have not yet been validated for AT1R. Finally, our results in humans must be seen as a first step. With a focus on safety, healthy volunteers were studied like in a clinical phase-1 trial. Given the significant regulatory burden for human use of a new imaging compound, this represents an important translational step and provides the foundation for subsequent disease-focused clinical studies.
In conclusion, targeted molecular imaging of cardiac AT1R expression is feasible using clinical PET/CT technology. In the future, this technique may provide unique insights into regional myocardial RAS activation in cardiac disease.
The authors thank the staff of the PET/CT center of Johns Hopkins Hospital for their excellent technical assistance.
This work was supported in part by NIH grant R01HL092985. Dr Fukushima was supported by a SNM Wagner-Torizuka fellowship. [11C]-KR31173 precursor was courtesy of Dr. Sung-Eun Yoo from Center for Biological Modulators/KRICT, Taejeon, Korea.
The authors report no conflicts of interest related to the subject matter of the article.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.