Author + information
- Received November 16, 2012
- Revision received March 9, 2013
- Accepted March 13, 2013
- Published online June 25, 2013.
- Thomas E. Ingram, MB, ChB, PhD∗,
- Alan G. Fraser, MB, ChB∗,
- Robert A. Bleasdale, MB, BCh†,
- Elizabeth A. Ellins, BSc, MA∗,
- Andrei D. Margulescu, MD∗,
- Julian P. Halcox, MA, MD∗ and
- Philip E. James, BSc, PhD∗∗ ()
- ∗Wales Heart Research Institute, Cardiff University Medical School, Cardiff, United Kingdom
- †Department of Cardiology, Royal Glamorgan Hospital, Ynysmaerdy, Llantrisant, Rhondda Cynon Taff, United Kingdom
- ↵∗Reprint requests and correspondence:
Dr. Philip E. James, Wales Heart Research Institute, Cardiff University Medical School, Heath Park, Cardiff CF 14 4XN, United Kingdom.
Objectives The aim of this study was to assess the potential benefits of inorganic nitrite in 2 clinical models: stress-induced myocardial ischemia and whole-arm ischemia-reperfusion.
Background Inorganic nitrite, traditionally considered a relatively inert metabolite of nitric oxide, may exert vasomodulatory and vasoprotective effects. Despite promising results from animal models, few have shown effectiveness in human model systems, and none have fully translated to the clinical setting.
Methods In 10 patients with inducible myocardial ischemia, saline and low-dose sodium nitrite (NaNO2) (1.5 μmol/min for 20 min) were administered in a double-blind fashion during dobutamine stress echocardiography, at separate visits and in a random order; long-axis myocardial function was quantified by peak systolic velocity (Vs) and strain rate (SR) responses. In 19 healthy subjects, flow-mediated dilation was assessed before and after whole-arm ischemia-reperfusion; nitrite was given before ischemia or during reperfusion.
Results Comparing saline and nitrite infusions, Vs and SR at peak dobutamine increased in regions exhibiting ischemia (Vs from 9.5 ± 0.5 cm/s to 12.4 ± 0.6 cm/s, SR from −2.0 ± 0.2 s−1 to −2.8 ± 0.3 s−1), whereas they did not change in normally functioning regions (Vs from 12.6 ± 0.4 cm/s to 12.6 ± 0.6 cm/s, SR from −2.6 ± 0.3 s−1 to −2.3 ± 0.1 s−1) (p < 0.001, analysis of variance). With NaNO2, the increment of Vs (normalized for increase in heart rate) increased only in poorly functioning myocardial regions (+122%, p < 0.001). Peak flow-mediated dilation decreased by 43% after ischemia-reperfusion when subjects received only saline (6.8 ± 0.7% vs. 3.9 ± 0.7%, p < 0.01); administration of NaNO2 before ischemia prevented this decrease in flow-mediated dilation (5.9 ± 0.7% vs. 5.2 ± 0.5%, p = NS), whereas administration during reperfusion did not.
Conclusions Low-dose NaNO2 improves functional responses in ischemic myocardium but has no effect on normal regions. Low-dose NaNO2 protects against vascular ischemia-reperfusion injury only when it is given before the onset of ischemia.
Inorganic nitrite (NO2−) was previously used extensively for angina (1–3). Because of variable timing and predictability of its effects, it fell out of favor when the faster acting and more potent organic nitrates became available (4). High NO2− doses can produce vascular collapse and syncope (5); we and others demonstrated recently that lower doses of sodium nitrite (NaNO2) mediate vasodilation preferentially during hypoxia (6,7). Thus, NaNO2 could exert clinically relevant anti-ischemic effects at much lower doses than previously used (8), while avoiding systemic adverse effects such as hypotension and tachyphylaxis (9).
Few recent clinical studies have investigated NO2−, and none have fully translated promising results from animal models. Beetroot juice contains high levels of inorganic nitrate (NO3−), which is bioconverted in the enterosalivary circuit to NO2−, increasing plasma [NO2−] (10). Beetroot concentrate enhances exercise performance in athletes (11,12), reduces blood pressure (by about 10 mm Hg), and attenuates ex vivo platelet activation in response to agonists (13). Beetroot juice protected against endothelial dysfunction induced by ischemia-reperfusion (IR) in the human forearm (14). Whether this was mediated by protective effects of NO2− during ischemia or reperfusion is unknown.
To determine the potential clinical value of low-dose intravenous NO2−, we investigated 2 different models of cardiovascular ischemic stress: first, inducible myocardial ischemia, in a double-blind, placebo-controlled study using dobutamine stress echocardiography (15,16), and second, forearm vascular IR injury, using flow-mediated dilation (FMD) to assess its pre-conditioning or post-conditioning effect.
All patients and subjects were refrained from alcohol, caffeine, and foods with high NO2− or NO3− content for 12 h before each study and fasted during the study. In our previous studies, this protocol minimized the influence of diet on baseline plasma [NO2−]. Plasma [NO2−] is similar in healthy control subjects and patients with stable coronary disease and is not influenced by age or fasting (Table 1).
The protocols conformed to the Declaration of Helsinki and were approved by the South East Wales Regional Ethics Committee. Each subject gave written informed consent.
The dose of NaNO2 given was the same as we have used previously (6). Either 1.5 μmol/min NaNO2 (Martindale Pharmaceuticals, Brentwood, United Kingdom) or 0.9% saline was administered by intravenous infusion for 20 min into an antecubital vein. This regimen produces a peak plasma [NO2−] level of about 350 nmol/l, approximately doubling basal concentrations (Fig. 1). Plasma NO3− and methemoglobin do not change significantly.
Study 1: myocardial ischemia
To test the hypothesis that NaNO2 improves regional function selectively in ischemic myocardial segments, we used myocardial velocity imaging to quantify regional myocardial function during dobutamine stress echocardiography in patients with known inducible myocardial ischemia.
Study patients had angina with ischemia documented by positive results on exercise tolerance tests (horizontal ST-segment depression >1 mm) and were awaiting revascularization for coronary artery disease confirmed by quantitative coronary arteriography (cross-sectional area stenosis >70%). Exclusion criteria were unstable symptoms requiring urgent treatment or ischemia on minimal exertion (positive at stage I or II of a Bruce exercise test).
A double-blind crossover study was performed (Fig. 2A). All subjects underwent 2 dobutamine stress echocardiographic examinations, separated by at least 7 days (mean interval: 10 days). They stopped all antianginal medications for 48 h before each study.
The protocol for each visit was identical, apart from the infusion. Either 1.5 μmol/min NaNO2 or 0.9% saline was started 10 min before the dobutamine infusion and continued for 20 min in total, achieving peak plasma NO2− during the peak dose of dobutamine.
Dobutamine Stress Echocardiography
A standard dobutamine stress echocardiographic protocol was used (17). Dobutamine was infused into a contralateral peripheral arm vein at 5, 10, 20, 30, and 40 μg/kg/min each for 3 min until the protocol was completed or the patient developed symptoms. Atropine was not used.
A commercially available echocardiography system (Vivid 7, GE, Fairfield, Connecticut) with a 2.5-MHz phased-array transducer was used to obtain cross-sectional grayscale and myocardial velocity images, with the subject in the left lateral decubitus position. Loops recorded during passively held end-expiration were stored digitally for post-processing using an EchoPAC workstation (GE). Each subject rested for 30 min before baseline recordings. At each stage, apical 4-chamber, 2-chamber, and long-axis images were recorded for 3 cardiac cycles. Image depth and width were optimized to result in frame rates >120 frames/s.
Regional myocardial function was quantified from digitally stored myocardial velocity loops, at the base of 6 left ventricular “walls”: basal septal and basal lateral segments in the apical 4-chamber view, basal anterior and basal inferior in the 2-chamber view, and basal anteroseptal and basal posterior in the long-axis view. Myocardial velocity was measured at the base and strain rate in the middle of each basal segment. Peak systolic velocity (Vs) was measured as the maximal Vs occurring after isovolumic contraction. Peak systolic strain rate was measured as the maximal strain rate during ventricular systole. All measurements were performed blinded to the study infusion and taken as the mean of 2 beats.
Quantification of Myocardial Responses
The percent increase of longitudinal Vs from baseline to peak-dose dobutamine (in subjects receiving saline) was calculated for each of the 6 basal wall segments in each subject. The response of a normal, nonischemic myocardial segment is an approximate doubling of Vs at peak stress (15); in this study, segments exhibiting the highest percent increases in Vs were taken to represent healthy myocardial segments, and segments exhibiting the lowest proportional increases in Vs were judged to demonstrate the greatest ischemic burden.
Study 2: vascular model of IR injury
To test the hypothesis that NO2− protects against ischemic vascular injury, as a pre-conditioning agent, we used an established model of vascular IR injury with FMD as a measure of endothelial function (18).
Healthy subjects underwent high-resolution ultrasonic assessment of brachial artery FMD before and 20 min after ipsilateral forearm ischemia created by inflating an upper-arm cuff to suprasystolic pressure.
On separate visits, each subject received an intravenous infusion into the contralateral arm (NaNO2 or 0.9% saline). This was given either before (pre-conditioning study) (Fig. 2B) or during (post-conditioning study) (Fig. 2C) whole-arm ischemia.
Endothelial function was assessed at baseline using FMD (19). Subjects lay supine in a temperature-controlled room. A B-mode scan of the brachial artery was obtained in the longitudinal axis, 5 to 10 cm above the antecubital fossa (Aloka ProSound SSD-5500; Hitachi Aloka, Tokyo, Japan). Electrocardiographically gated end-diastolic images were acquired every 3 seconds for off-line analysis. A 1-cm to 2-cm segment of artery was selected for measurement of diameter, using automated edge detection software (Brachial Analyser for Research 5, Medical Imaging Applications LLC, Coralville, Iowa). Blood velocity recorded continuously at the center of the brachial artery using pulsed Doppler was analyzed using the same software.
Arterial flow was manipulated by a cuff placed around the forearm immediately below the antecubital fossa. After 1 min of baseline recording, the cuff was inflated to 250 mm Hg for 5 min and then released, resulting in reactive hyperemia. Arterial images were recorded for a further 5 min. Maximal arterial diameter was determined and change from baseline calculated and expressed as a percent age.
Transient whole-arm endothelial dysfunction was then produced by inflating another cuff around the upper arm to 250 mm Hg for 20 min. FMD was reassessed 20 min after release of this cuff, with the difference from baseline measurements representing the degree of injury resulting from the IR stimulus.
These clinical proof-of-concept investigations were powered on the basis of published studies. In our laboratory, the coefficient of variation for repeated measurements of myocardial Vs is about 10% (20), implying that in 10 subjects, it should be possible to demonstrate changes of 9% between repeated measurements at p < 0.05. In 10 subjects, FMD demonstrated that beetroot juice prevented endothelial dysfunction induced by a similar IR protocol at similar plasma concentrations of NO2−(14).
Increments of myocardial velocity during dobutamine treatment are influenced by heart rate (15). We therefore compared absolute peak responses and the rate of increase in Vs adjusted for changes in heart rate.
Changes in longitudinal velocity in basal segments are sensitive to the development of ischemia in any myocardial segment (15). Measurements of velocity are more reproducible than measurements of deformation (strain rate or peak systolic strain). The mean percent increase in Vs in ischemic segments at the peak dose of dobutamine is ≤50% to 75%, compared with about 100% in nonischemic segments, but there is a continuum between ischemic and healthy responses because of tethering between contiguous segments. In this study, 6 basal segments were analyzed per patient per visit; the 2 segments with the lowest responses (in 89% of instances within or adjacent to the territory of a coronary artery with a significant stenosis) were compared with the 2 segments with the highest responses (in the territory of coronary arteries without angiographic stenosis). The lower tertile represented the ischemia group and the upper tertile the normal group.
Where appropriate, linear regression analysis was performed comparing data from the pre-specified groups at different stages of the protocol. The linear regression gradient for each subject in each group was calculated. The resulting mean linear regression gradients were compared using repeated-measures analysis of variance with Newman-Keuls post-tests, unless only 2 groups of data were present, in which case paired t tests were used. Continuous variables are reported as mean ± SEM. Analyses were performed using GraphPad Prism version 4.0 (GraphPad Software, La Jolla, California).
Study 1: myocardial ischemia
Patient characteristics are given in Table 2. There were no differences in heart rate or blood pressure between groups (p = NS) (details in Online Appendix). Most dobutamine stress echocardiographic studies were stopped because of the onset of typical angina; the full protocol was completed in 5 studies. No significant adverse events occurred.
Velocity profiles from 55 of 60 myocardial walls (6 walls in each of 10 subjects) were of sufficient quality for analysis of Vs throughout the study. Thus, the ischemic and control groups each contained data from 18 basal segments.
In normal walls, Vs increased in response to dobutamine, similarly during saline (+117%) and NaNO2 (+122%) (p = NS) (Table 3, Fig. 3). In ischemic walls, there was an attenuated increase during saline infusion (+51%) but a normal increase during NaNO2 infusion (+132%). Similar changes were observed in longitudinal systolic strain rate, which increased by 79% and 66% in normal segments in response to dobutamine during saline and NaNO2 infusions, respectively, whereas in the abnormal segments, it increased by only 26% during saline compared with 118% during NaNO2.
The change in Vs from baseline to peak dobutamine in ischemic segments, adjusted for change in heart rate, was +3.7 ± 0.6 cm/s/s during the control intervention (ischemia/saline) and +8.2 ± 1.0 cm/s/s during NaNO2 infusion (ischemia/NO2−) (p < 0.01 vs. baseline, comparable with normal segments). Responses to dobutamine stress echocardiography in the healthy segments were not influenced by NaNO2 (normal/saline: +10.5 ± 1.1 cm/s/s; normal/NO2−: +8.4 ± 0.7 cm/s/s) (p = NS, but p < 0.01 for each compared with the ischemia/saline group) (Fig. 3).
The mean linear regression gradient for the increase in Vs relative to the increase in heart rate was 0.085 ± 0.017 in the ischemia/saline group. This was lower (p < 0.01) than in the other groups (ischemia/NO2−: 0.132 ± 0.015; normal/saline: 0.164 ± 0.019; normal/NO2−: 0.140 ± 0.011) (Online Appendix).
Study 2: IR injury
The pre-conditioning protocol was performed in 10 subjects (6 men) age 24.6 ± 1.4 years, with a mean body mass index of 23.7 ± 0.9 kg/m2. When saline was infused before the initiation of ischemia, peak FMD was reduced on average by 42.6% after vascular ischemia (FMD 1 = 6.8 ± 0.7% vs. FMD 2 = 3.9 ± 0.7%; p < 0.01), confirming that the stimulus induced endothelial dysfunction (Table 4, Fig. 4). The IR injury was prevented when NaNO2 infusion was delivered before the initiation of ischemia (FMD 1 = 5.9 ± 0.7% vs. FMD 2 = 5.2 ± 0.5%; p = NS). There was no difference in baseline endothelial function (FMD 1) between the two visits.
The post-conditioning protocol was performed in 9 subjects (5 men) age 27.7 ± 2.0 years, with a mean body mass index of 24.3 ± 1.3 kg/m2. In this protocol, each infusion was started after initiation of ischemia (i.e., it was delivered to the ischemic tissue during reperfusion). There was a mean reduction in FMD of 35% after IR when saline was administered (from FMD 1 = 8.6 ± 0.6% to FMD 2 = 5.6 ± 1.0%; p < 0.05). The decrease in FMD after IR was similar (−41%) when NaNO2 was infused (from FMD 1 = 7.5 ± 0.5% to FMD 2 = 4.4 ± 0.5%; p < 0.05) (Table 4).
We demonstrated in 2 complementary clinical settings that low-dose NaNO2 reduces the effects of ischemia, ameliorating inducible myocardial ischemia and protecting against IR-mediated vascular injury, as long as it is administered before the onset of ischemia.
In both clinical models, low-dose NO2− was effective when given intravenously for only 20 min, and its therapeutic benefits were demonstrable within 5 to 20 min. Local effects were observed, without the systemic side effects associated with organic nitrates and other nitric oxide (NO) donors. Together, these studies imply that low-dose NaNO2 is most useful if it is given before elective procedures such as coronary artery bypass surgery, percutaneous coronary intervention, and organ transplantation. It is less likely to be beneficial if initiated after the onset of ischemia.
NaNO2 as a targeted vasodilator
NaNO2 is a weak vasodilator in normoxia, compared with other NO donors such as glyceryl trinitrate (21). Previously, large doses of NaNO2 given to patients with angina to achieve a quick therapeutic response often produced profound and sustained systemic vasodilation.
Our results demonstrate that low-dose NaNO2 can reduce inducible myocardial ischemia in patients with stable obstructive coronary artery disease, without influencing normoxic vasculature. NaNO2 appears to act as a targeted coronary vasodilator, providing an NO-like effect only in ischemic regions. NO2− exhibits vasoactivity mostly by reduction to NO, via pathways that are exacerbated in hypoxia (7,22); the principal candidates are deoxyhemoglobin (8), xanthine oxidase (22), myoglobin (23), and aldehyde oxidase (24,25). It remains unclear whether NO2− reduction occurs primarily in blood or tissue. Whether NO2− acts selectively on conductance or resistance vessels is also unclear and likely to be a balance of pO2/HbO2 saturation, decreased pH, and reductive state of oxidoreductase enzyme systems. NO2− also exhibits acute and direct vasodilation, most likely through cyclooxygenase-mediated pathways (8).
Long-term low-dose NaNO2 supplementation may be an effective alternative treatment for angina, given orally either as NaNO2 itself, or as large doses of inorganic NO3−, which is converted by oral bacteria to NO2−(26,27). Beetroot concentrate, high in NO3−, can increase plasma NO2− to about 800 nmol/l; much lower dietary intake of beetroot concentrate elevates plasma NO2− modestly, to levels similar to those attained in our subjects. Importantly, plasma NO2− remains elevated for several hours after a single loading dose. Organic nitrates are likely to remain the primary pharmacologic means of achieving a rapid NO-mediated effect in clinical practice, but long-term NaNO2 therapy might prevent tolerance (28). Primates given a continuous intravenous infusion of NaNO2 for 14 days did not exhibit tachyphylaxis (29).
Our studies were not powered to address the influence of diet and resulting plasma levels on acute responses to NaNO2. However, we have not observed large variations in plasma NO2− across several populations unless it was consumed during the previous 12 h. Plasma NO3− varies considerably more, and dietary consumption of large amounts of NO3− does have a direct influence on plasma NO2− level, and it can affect blood pressure and platelet responsiveness to agonists (10,13,14,26,27). Our subjects were not fasting, but they refrained from foods high in NO2− or NO3−. The question remains whether subjects with higher basal levels will also benefit from the modest elevation in NO2− achieved by the infusion we gave (to about 200 nmol/l).
Effects of NO2− supplementation
Animal studies have demonstrated a protective effect of long-term oral NaNO2 supplementation during experimental myocardial infarction (30,31). The cardiovascular benefits of a “healthy” Mediterranean diet and certain Chinese medicines could be partly explained by an increase in NO2−(32,33). Whether NaNO2 can improve clinical outcomes requires evaluation in adequately powered, randomized clinical trials. It is possible that long-term NaNO2 treatment could reduce clinical end points where long-term augmentation of the endothelial NO synthase–NO axis (with L-arginine) or long-term NO donors (organic nitrates) has not. Uncoupling of endothelial NO synthase and NO donors can produce other metabolites, including the potentially harmful free radical peroxynitrite (34). The efficacy of NO donors depends on a healthy endothelium and endothelial microenvironment, but endothelial function deteriorates with age and increasing risk factors. Administration of NaNO2 in this setting benefited patients with peripheral vascular diseases (35).
NaNO2 as a pre-conditioning but not a post-conditioning agent
We showed that NaNO2 administered before an ischemic insult protects against inducible endothelial dysfunction after reperfusion. There was no protective effect when administered after the onset of ischemia, such that higher concentrations of NaNO2 were delivered only during reperfusion.
Our results contradict those of Gonzalez et al. (36), who reported that NaNO2 provided effective post-conditioning in a canine acute myocardial infarction model. Myocardial IR injury was reduced when high doses of NaNO2 were given systemically during the final 5 min of a 60-min period of ligation of a coronary artery. This intervention increased plasma NO2− 30-fold (to 6 μmol/l), greatly exceeding plasma levels achieved in our studies. Alternative explanations could be that such high levels lead to direct absorption into ischemic myocardium from the left ventricular cavity, adjacent perfused myocardium, or small collateral vessels, activating protective pathways before reperfusion. An intraventricular injection of NaNO2 given immediately before reperfusion protected against injury in rats (37), but most animal models have demonstrated benefit when NaNO2 was injected before ischemia (38–40).
Our “proof-of-concept” studies provide limited mechanistic insights into how NaNO2 affords cytoprotection against ischemic injury. Several potential pathways are summarized in Figure 5. Most benefits of NaNO2 are thought to derive from its conversion to NO; activation of soluble guanylate cyclase or protein kinase and mitochondrial adenosine triphosphate–sensitive potassium channels limit adenosine triphosphate depletion and mitochondrial calcium uptake (41). Further, S-nitrosation of key intracellular proteins may reduce mitochondrial L-type Ca2+ channel function (42) and inhibit complex 1 of the electron transport chain, which generates reactive oxygen species during reperfusion (43,44). At the modest and near-physiological doses of NaNO2 we used, we cannot confirm whether nitroglycerin-based and/or NO-based cytoprotection observed at higher doses is also relevant.
We demonstrated previously that the vasoactivity of NaNO2 during hypoxia does not depend on concurrent elevation of plasma NO2−. After similar low doses in healthy subjects, hypoxic pulmonary vasoconstriction was alleviated for several hours despite plasma NO2− returning to near-baseline values (6). This implies that NO2− enters the tissue and can be used during hypoxia; or it is converted to a metabolite that is stored in tissue; or its effect can be transmitted from the surface of the vessel into tissue, where it prevents subsequent insult.
Low-dose NaNO2 reduced the effects of myocardial ischemia in patients with stable coronary artery disease and reduced vascular injury in healthy subjects when administered before an IR insult. These effects occurred without systemic vasodilation, suggesting that the benefits of low-dose NaNO2 are selectively targeted to ischemic myocardium and vascular tissue. Thus, what has been considered historically to be the weakest pharmacokinetic property of NO2− could turn out to be its greatest asset.
The authors thank Catherine Templeton (Royal Glamorgan Hospital) for assistance in collecting physiological parameters during echocardiographic studies.
This work was supported by the British Heart Foundation (FS/06/088). The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Fraser and James contributed equally to this work in terms of senior authorship.
- Abbreviations and Acronyms
- flow-mediated dilation
- sodium nitrite
- nitric oxide
- systolic velocity
- Received November 16, 2012.
- Revision received March 9, 2013.
- Accepted March 13, 2013.
- American College of Cardiology Foundation
- Butler A.
- Reichart E.,
- Mitchell S.
- Ingram T.E.,
- Pinder A.G.,
- Bailey D.M.,
- Fraser A.G.,
- James P.E.
- Maher A.R.,
- Milsom A.B.,
- Gunaruwan P.,
- et al.
- Gladwin M.T.,
- Shiva S.,
- Dezfulian C.,
- Hogg N.,
- Kim-Shapiro D.B.,
- Patel R.P.
- ↵(2007) British National Formulary (BMJ Publishing Group, London, United Kingdom), 54th ed.
- Kapil V.,
- Webb A.J.,
- Ahluwalia A.
- Mädler C.F.,
- Payne N.,
- Wilkenshoff U.,
- et al.
- Voigt J.U.,
- Exner B.,
- Schmiedehausen K.,
- et al.
- Sicari R.,
- Nihoyannopoulos P.,
- Evangelista A.,
- et al.
- Loukogeorgakis S.P.,
- Panagiotidou A.T.,
- Broadhead M.W.,
- Donald A.,
- Deanfield J.E.,
- MacAllister R.J.
- Ignarro L.J.,
- Lippton H.,
- Edwards J.C.,
- et al.
- Godber B.L.,
- Doel J.J.,
- Sapkota G.P.,
- et al.
- Ormerod J.O.,
- Ashrafian H.,
- Maher A.R.,
- et al.
- Li H.,
- Cui H.,
- Kundu T.K.,
- Alzawahra W.,
- Zweier J.L.
- Carlström M.,
- Persson A.E.,
- Larsson E.,
- et al.
- Butler A.R.,
- Feelisch M.
- Dejam A.,
- Hunter C.J.,
- Tremonti C.,
- et al.
- Bryan N.S.,
- Calvert J.W.,
- Elrod J.W.,
- Gundewar S.,
- Ji S.Y.,
- Lefer D.J.
- Pattillo C.B.,
- Bir S.,
- Rajaram V.,
- Kevil C.G.
- Gonzalez F.M.,
- Shiva S.,
- Vincent P.S.,
- et al.
- Zhao Z.-Q.,
- Corvera J.S.,
- Halkos M.E.,
- et al.
- Webb A.,
- Bond R.,
- McLean P.,
- Uppal R.,
- Benjamin N.,
- Ahluwalia A.
- Heusch G.,
- Boengler K.,
- Schulz R.
- Sun J.,
- Morgan M.,
- Shen R.-F.,
- Steenbergen C.,
- Murphy E.
- Rogers S.C.,
- Khalatbari A.,
- Gapper P.W.,
- Frenneaux M.P.,
- James P.E.