Author + information
- Received January 30, 2004
- Revision received February 27, 2004
- Accepted March 2, 2004
- Published online August 4, 2004.
- Stefano Fiorucci, MD⁎,
- Andrea Mencarelli, BS⁎,
- Alessandra Meneguzzi, BS†,
- Alessandro Lechi, MD†,
- Barbara Renga, BS⁎,
- Piero del Soldato, PhD‡,§,
- Antonio Morelli, MD⁎ and
- Pietro Minuz, MD†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Pietro Minuz, Medicina Interna C, Policlinico GB Rossi, Piazzale LA Scuro, 10, 37134 Verona, Italy.
Objectives The goal of this study was to test the hypothesis that NCX-4016 may have broader anti-inflammatory and antithrombotic effects as well as better gastric tolerability than aspirin in humans.
Background NCX-4016 is an aspirin derivative containing a nitric oxide-releasing moiety that prevents platelet activation and modulates tissue factor (TF) expression and cytokine release from lipopolysaccharide (LPS)-stimulated monocytes.
Methods This was a blind-observer, placebo-controlled, parallel-group study in which 48 healthy subjects were randomized to receive NCX-4016 800 mg twice a day, NCX-4016 800 mg twice a day plus aspirin 325 mg, aspirin 325 mg, or placebo for 21 days.
Results Similar to aspirin alone, NCX-4016 effectively inhibited platelet aggregation induced by 0.6 mmol/l arachidonic acid, clot-stimulated thromboxane (TX) B2generation in whole blood, and urinary excretion of 11-dehydro-TXB2. Unlike aspirin alone, the administration of NCX-4016 significantly inhibited TF expression in monocytes stimulated ex vivo with 10 μmol/l LPS (determined by flow-cytometry analysis of TF on CD14 positive cells). NCX-4016 also inhibited the rapid TF expression induced in monocytes by a proteinase activated receptor agonist (thrombin receptor activator protein, 2 μmol/l) as well as LPS-induced expression of CD11b . Ex vivo, release of MCP-1 and interleukin-6 were significantly inhibited by NCX-4016, but not by aspirin. NCX-4016 was not associated with gastric damage, and significantly reduced gastric injury when co-administered with aspirin, although both drugs reduced gastric PGE2production to the same extent.
Conclusions NCX-4016 is equally effective as aspirin in inhibiting cyclooxygenase activity. However, NCX-4016 causes less gastric damage and prevents monocyte activation. Larger multicenter trials are warranted to establish clinical efficacy and safety of NCX-4016.
NCX-4016 (2-[acetyloxy]benzoic acid 3-[nitrooxymethyl]phenyl ester) is an acetylsalicylic acid (aspirin) (ASA) derivative containing a nitric oxide (NO)-releasing moiety that has been investigated as an antithrombotic and anti-inflammatory drug (1). NCX-4016 inhibits cyclooxygenase (COX) activity in platelets both in vitro and ex vivo and, similar to ASA, prevents the release of thromboxane (TX) A2and arachidonic acid-induced platelet aggregation (2–4). NCX-4016 has also been found to inhibit ASA-insensitive platelet aggregation as well as platelet adhesion and the expression of platelet adhesion molecules in vitro, unlike its parent compound (1). These activities were prevented by the use of NO scavengers, indicating that at least part of the antiplatelet activity of NCX-4016 is COX-independent and mediated by NO (2).
We hypothesized that, in comparison with ASA, NCX-4016 might have anti-inflammatory activities that are related to its antithrombotic effects as inflammation and thrombosis are tightly linked (5,6). Tissue factor (TF) is the major determinant of thrombogenicity of atherosclerotic plaques (7–9). Monocytes expressing TF and MAC-1 (CD11b) and proinflammatory cytokines and chemokines exert a mechanistic role in the complex choreography of atherothrombosis ranging from the induction of a prothrombotic phenotype in endothelial cells to the recruitment of leukocytes and the generation of unstable atherosclerotic plaques (6,10).
While investigating the effect of NCX-4016 in lipopolysaccharide (LPS)-stimulated human monocytes, we found a reduction in the cellular release of cytokines as well as the expression of TF in response to NCX-4016, but not with ASA (11,12). Similarly, in a rodent model of inflammation, in vivo administration of NCX-4016, but not ASA, decreased TF expression and activity on circulating monocytes (13), suggesting that this compound exerts NO-mediated, ASA-independent, anti-inflammatory activities.
The present study evaluated whether administration of NCX-4016 prevents activation of both platelets and monocytes and also assessed its gastric tolerability. NCX-4016 was given to healthy subjects for a prolonged period of time and compared with ASA, administered alone or in association with NCX-4016. By using proteinase activated receptor (PAR)-1 agonist and LPS as stimuli, we also explored the mechanism by which NCX-4016 modulates TF expression. Finally, by examining the combined effects of NCX-4016 and ASA on gastric mucosa and platelet and monocyte functions, we demonstrated an increased gastric tolerability and potential antithrombotic effects of NCX-4016.
Forty-eight healthy subjects (32 men and 16 women; range, 19 to 44 years) were enrolled. Exclusion criteria were: gastrointestinal symptoms; alcohol or drug abuse; pregnancy; intolerance or hypersensitivity to ASA; a history of liver, kidney, or cardiac disease; peptic ulceration; or upper digestive tract surgery. Subjects were also excluded if they had received any medication in the three weeks before the study, including herbal remedies. Subjects were excluded if they had abnormal findings on baseline endoscopy (total endoscopic score >1) or an abnormal baseline biochemical profile. A serum sample from each subject was analyzed for the presence of IgG antibodies to Helicobacter pylori(EIAgen kit, Biochem Immunosystem, Bologna, Italy).
This was a blind-observer, parallel-group study in which subjects were randomized into four groups composed of 12 subjects each that received one of the following treatments: 1) NCX-4016 placebo twice daily; 2) NCX-4016 800 mg twice daily; 3) NCX-4016 800 mg twice daily plus commercial ASA 325 mg once daily; 4) commercial ASA 325 mg once daily plus NCX-4016 placebo once daily. Randomization was performed in blocks of four from a computer-generated list. The medications were taken orally at 8 amand 8 pmat the clinical site for 21 consecutive days. On the mornings of the first (day 0) and the last day of the study (day 22), approximately 12 h after the last drug administration, endoscopic examination was performed. Systolic and diastolic blood pressures were measured in the supine and standing position before, 1, 3, and 6 h after drug administration. Heart rate was also recorded. Blood samples were collected immediately before and 4 h after drug administration for basal and post-treatment biochemical analyses, platelet aggregation and serum TXB2studies. The urinary excretion of 11-dh-TXB2was assessed on days 0 and 21 (6-h urine collections, from 7 amto 1 pm). Tissue factor expression on blood monocytes and the ex vivo release of MCP-1 and interleukin (IL)-6 were determined on day 21, 4 h after the last drug administration. The study was approved by local ethics committee, and written informed consent was obtained from all subjects.
Platelet aggregation was assayed according to Born’s turbidimetric method using a 4-channel aggregometer (APACT4, Helena Biosciences Europe, Sunderland, United Kingdom). Blood was collected and anticoagulated with 3.8% sodium citrate, and platelet-rich plasma was obtained by centrifugation at 200 × gfor 10 min. Platelet-rich plasma was incubated for 3 min at 37°C, and one of the following agonists was then added: arachidonic acid (0.6 mmol/l), the PAR-1 agonist SFLLRNPNDKYEPF (thrombin receptor activator peptide [TRAP], Sigma Chemical Co., St. Louis, Missouri; 10 μmol/l), the TXA2analogue U46619 (1.5 μmol/l). The percentage of aggregation was determined 3 min after the addition of platelet agonists.
Gastroduodenal endoscopy and mucosal biopsies
Upper endoscopies were performed using an Olympus Exera GIF-Q160 apparatus (Olympus Europe, Hamburg, Germany). The entire stomach and duodenum were systematically examined from the fundus to the duodenum. Endoscopies were evaluated by an independent gastroenterologist and by a medical advisor for gastroscopy under blind conditions. Each endoscopic procedure was completely recorded using a VHS recorder. Mucosal lesions were graded according to a modified Lanza’s scale: grade 0 = normal mucosa; grade 1 = 1 to 3 erosions or submucosal hemorrhages; grade 2 = 4 to 10 erosions or submucosal hemorrhages; grade 3 = >10 erosions or submucosal hemorrhages; grade 4 = ulcer or diffuse submucosal hemorrhages. Separate endoscopic injury scores were assigned for the esophagus, gastric fundus, gastric body, gastric antrum, and duodenum (4).
At the end of endoscopic evaluation, four biopsies were taken from the antrum. Biopsy specimens were frozen in liquid nitrogen and then added to a tube containing 100% ethanol plus 0.1 mmol/l indomethacin and were homogenized and centrifuged for 10 min at 12,000 rpm at 4°C. The supernatant of each sample was analyzed for the amount of PGE2and PGF2α, major prostaglandins synthesized in the human gastric mucosa, using commercial enzymatic immunoassay (EIA) kit (PGE2and PGF2αEIA kits, Cayman Chemical Co., Ann Arbor, Michigan).
For the measurement of serum TXB2, the stable non-enzymatic metabolite of TXA2, 5 ml of venous non-anticoagulated blood samples were allowed to clot for 1 h at 37°C in glass tubes and then centrifuged at 3,000 gfor 15 min. Serum samples were then stored at −80°C until assayed by EIA (Cayman Chemical Co.).
Urinary TXA2-metabolite excretion
Urine samples (40 ml) were centrifuged and stored at −20°C until chromatographic purification procedures were performed. 11-dehydro-TXB2, the major TXA2metabolite in urine, was then measured by EIA using a specific monoclonal antibody (Cayman Chemical Co.).
Flow-cytometry analysis of TF expression
Blood samples (3 ml) were collected using either sodium citrate or sodium heparin as anticoagulants; TRAP (2 μmol/l) was added to citrate-anticoagulated blood samples, which were then incubated for 15 min at 37°C. Bacterial LPS (Escherichia coli, O113:H10, Sigma Chemical Co.) was added to heparin-anticoagulated blood samples and incubated for 6 h. After incubation with agonists, blood samples were diluted in lysis buffer for red cells (155 mmol/l NH4Cl, 10 mmol/l KHCO3, and 0.1 mmol/l EDTA). White blood cells were then incubated with anti-CD14 PC5-conjugated monoclonal antibody for 30 min at 4°C (Beckman Coulter, Fullerton, California), washed, and stained with goat anti-human FITC-conjugated anti-TF monoclonal antibody (American Diagnostica, Stamford, Connecticut) for 30 min at 4°C. For CD11b staining, cells were incubated with FITC-conjugated goat anti-human CD11b monoclonal antibody (Beckman Coulter). Freshly stained cells were analyzed by double-color fluorescence distribution on an EPIX XL cytofluorimeter (Beckman Coulter) and the number of CD14-TF positive cells counted; the mean intensity of florescence was also calculated.
MCP-1 and IL-6 secretion
Plasma obtained from heparin-anticoagulated and LPS-stimulated blood samples was assayed for MCP-1 and IL-6 concentration using specific enzyme immunoassays (Endogen Pierce Biotechnology, Rockford, Illinois).
The calculation of the sample size and power for the analysis of variance (ANOVA) with contrasts, was made with PASS 2002 software (NCSS Statistical Software, Kaysville, Utah). The total sample of 48 subjects achieves about 85% power to detect a difference of at least 5 in the mucosal damage scores. The common SD within a group was assumed to be 2. An alpha level of 0.05 was used for all statistical tests. Data were analyzed using the SAS/STAT version 8.2 (SAS Institute Inc., Cary, North Carolina). The primary objectives of the study were to compare the mucosal damage scores, platelet aggregation, and TXB2at day 21. Differences between treatment groups were tested using the ANOVA followed by Tukey-Kramer test for pairwise multiple comparisons. Values of p < 0.05 were considered significant. Data are expressed as mean ± SE.
All volunteers completed the study protocol. There were no statistically significant differences between the basal and end-of-treatment hematic-biochemical values in any of the four groups (data not shown). No statistically significant differences in supine and standing blood pressure or heart rate were observed between subjects receiving placebo or any of the active treatments at baseline or at days 1, 7, and 21 (data not shown). Serum IgG antibodies to H. pyloriwere found in 23 subjects: 6 in the placebo group, 6 in the ASA-treated group, 4 in the NCX-4016-treated group, and 7 in NCX-4016 plus ASA group (p not significant among groups).
Effect of ASA and NCX-4016 on gastrointestinal mucosa
No visible injury was present in either the stomach or duodenum in any of the 48 participants at baseline. As shown in Figure 1A,three weeks administration of ASA 325 mg once daily resulted in appearance of gastric and/or duodenal lesions with a mean total endoscopic score of 13.3 ± 2.1 (p < 0.0001 vs. placebo). Administration of 800 mg twice a day NCX-4016 caused only minimal gastrointestinal injury, similar to that observed in the placebo group, with a mean total endoscopic score of 1.8 ± 0.5 (1.7 ± 0.7 in the placebo group; p < 0.0001 vs. ASA). Co-administration of NCX-4016 and ASA attenuated the gastroduodenal damage caused by ASA, resulting in a mean endoscopic score of 4.8 ± 0.8 (p < 0.001 vs. ASA). Administration of NCX-4016 and the combination of the two caused an 80% reduction in gastric PGE2and PGF2αconcentrations (Fig. 1B) (p < 0.001 vs. placebo).
Effect of ASA and NCX-4016 on platelet aggregation
No differences in arachidonic acid-induced aggregation were observed among treatment groups under basal conditions. Platelet aggregation was not induced in subjects administered NCX-4016, ASA, or both drugs for 21 days, either immediately before or 4 h after the last drug administration (Fig. 2A).Aggregation induced by thrombin or U46619 was not affected by any treatment (data not shown).
Effect of ASA and NCX-4016 on serum TXB2and TXA2-metabolite excretion
Both ASA and NCX-4016 reduced the serum concentrations of TXB2by more than 90% (p < 0.0001 vs. placebo), although ASA was more effective (p < 0.001 vs. NCX-4016). Co-administration of NCX-4016 and ASA did not interfere with the antiplatelet activity of the latter drug (Fig. 2B). Aspirin and NCX-4016 were also equally effective in reducing the urinary excretion of 11-dehydro-TXB2(Fig. 2C). Again, co-administration of NCX-4016 and ASA did not affect the inhibitory activity of either drug.
Effect of ASA and NCX-4016 on TF expression and cytokine/chemokine generation in ex vivo-stimulated monocytes
As illustrated in Figure 3,treatment of CD14+ cells with TRAP and LPS resulted in a significant upregulation of TF expression, as measured by assessing the percent of TF-positive monocytes (Figs. 3A and 3B) and mean intensity fluorescence (Figs. 3C and 3D, n = 12, p < 0.01 vs. untreated cells). Administration of NCX-4016 resulted in significant attenuation of ex vivo upregulation of TF induced by either TRAP or LPS (p < 0.05 vs. agonists). NCX-4016 also maintained this effect when co-administered in combination with ASA. However, ASA alone failed to modulate the expression of TF.
Ex vivo exposure of CD14-positive cells to LPS resulted in significant upregulation of both MCP-1 and IL-6 (Figs. 4Aand 4B, p < 0.01 vs. untreated). Lipopolysaccharide stimulation of blood samples obtained from volunteers treated with NCX-4016 resulted in a 50% reduction of MCP-1 and a 25% reduction of IL-6 (p < 0.05 vs. placebo and ASA). This effect was maintained in volunteers treated with NCX-4016 plus ASA (p < 0.05 vs. placebo and ASA alone), but not in volunteers treated with ASA alone, indicating that the NO-releasing moiety of NCX-4016 was likely responsible for these effects. Administration of NCX-4016 resulted in significant reduction of CD11b expression on CD14-positive cells (Figs. 4C and 4D) (p < 0.05 vs. placebo). Moreover, this effect was maintained in volunteers administered with NCX-4016 in combination with ASA, while ASA by itself was ineffective (Figs. 4C and 4D).
NCX-4016 contains two pharmacologically active moieties, namely an ASA-like moiety and a NO-releasing moiety, which both contribute to its pharmacologic effects. In fact, this compound has dual pharmacodynamic effects, acting as a COX inhibitor and a NO donor. In the present study, we explored the capacity of NCX-4016 to reduce the generation of TXA2in platelets by acting as a COX inhibitor. This has been previously shown in vitro and ex vivo both in experimental animals and humans by testing the release of TXB2from aggregating platelets induced by blood coagulation (1,2,4). Herein, we demonstrate that both NCX-4016 and ASA reduce the clot-stimulated TXB2release from aggregating platelets by more than 90%. This inhibitory effect is associated with both the prevention of arachidonic acid-induced platelet aggregation and substantial reduction in the urinary excretion of 11-dehydro-TXB2, thus suggesting that NCX-4016 satisfies the requirement of a sustained inhibition in a platelet synthetic capacity that was shown to be necessary for a COX inhibitor in order to prevent in vivo platelet activation (14,15). The urinary excretion of 11-dehydro-TXB2represents a reliable index of in vivo platelet activation and predicts adverse cardiovascular events in high-risk patients (16). Oral administration of ASA to healthy subjects reduces the urinary excretion of 11-dehydro-TXB2by about 80%. Extraplatelet sources may account for the residual excretion of 11-dehydro-TXB2, which increases under clinical and experimental conditions related to atherothrombosis due to the contribution of COX-2-expressing inflammatory cells (17). Although NCX-4016 has been shown to inhibit both platelet COX-1 activity and COX-2 expression in monocytes (13), we may infer that platelet COX is the target of both drugs in healthy subjects, because no additive effect on the urinary excretion of 11-dehydro-TXB2is observed when NCX-4016 and ASA were co-administered. The present results are consistent with the previous demonstration that ASA is actually released in vivo from NCX-4016 and is, furthermore, responsible for its ASA-like activity (18).
If both ASA and NCX-4016 effectively inhibit platelet COX, the antithrombotic activity of both drugs should be expected to be similar. However, previous data indicate that NCX-4016 is superior to ASA in an animal model of platelet-mediated pulmonary embolism (19), suggesting that NCX-4016 has in vivo COX-independent, NO-mediated properties that are not shared by ASA, such as the anti-adhesive effects shown in vitro (2). Consistent with this view, we have previously shown that NCX-4016, but not ASA, inhibits TF expression on LPS-stimulated monocytes (11,12). Both synthesis and expression of TF, as well as TF-dependent procoagulant activity and TXA2biosynthesis, were reduced by NCX-4016, further supporting the concept that both COX-dependent and -independent mechanisms are responsible for the action of this drug (13,20). Therefore, we tested the hypothesis that NCX-4016 has effects on monocyte function in humans by investigating its effects on TF expression.
Tissue factor expression induced by LPS occurs primarily via gene transcription and is a delayed effect (21). Experimental evidence supports the hypothesis that NO is responsible for downregulation of TF expression acting on protein synthesis (22). However, when NCX-4016 and nitrosorbide mononitrate was tested in LPS-challenged animals, the expression of TF on monocytes was also downregulated in the absence of a reduction in TF messenger RNA and protein synthesis, suggesting that NO can also modulate TF expression by altering its intracellular traffic by reducing the assembly of preformed TF on the outer side of the cell membrane (13). The data from the present study give additional weight to this hypothesis (13,22), and, in fact, we have demonstrated that surface expression of TF on monocytes can be induced within minutes via stimulation of PAR-1 without requiring messenger RNA synthesis (23); unlike ASA, NCX-4016 almost completely prevents PAR-1-mediated surface expression of TF.
While direct modulation of TF expression on CD14+ cells might well explain the effect of NCX-4016, there is evidence that TF expression is amplified by the interaction of platelets with monocytes (23,24). This interaction requires the ligation of platelet P selectin and the release of soluble factors (25). Because PAR-1 is expressed on both platelets and monocytes and NCX-4016 acts on both cell types, it cannot be excluded that TRAP-induced monocyte activation is also modulated by the antiplatelet activities of NCX-4016 (2).
Another important observation of this study is the demonstration that IL-6 and MCP-1 secretion as well as CD-11b expression on circulating monocytes was reduced in subjects receiving NCX-4016, which may be of interest when the role of inflammation in vascular disease is considered. Interleukin-6 and MCP-1 are known amplifiers of atherogenic changes and restenosis by inducing the migration of circulating monocytes and activating resident monocytes/macrophages and by promoting changes in endothelial and smooth muscle cells (26,27). Because MCP-1 is a stimulus for TF expression on smooth muscle cells (28), it might be hypothesized that NCX-4016 is able to interrupt the amplificatory loops in the inflammatory/thrombotic process centered on TF expression. None of these anti-inflammatory effects is observed with ASA unless NCX-4016 is co-administered. Along these lines, the NO-mediated inhibition of transcription factors involved in the inflammatory response might explain the multiple anti-inflammatory effects shown by NCX-4106 in animals (29,30). However, post-translational regulation of the proinflammatory machinery could also be involved, because NCX-4016 has already been shown able to modulate IL-1β release, through a NO-mediated inhibition of caspase-1 interleukin converting enzyme activity (12).
Several lines of evidence indicate that the administration of NCX-4016 is associated with less gastric damage than ASA (4). The gastric tolerability of NCX-4016 (800 mg twice a day) has already been shown to be superior to that of ASA (100 mg once a day) (31). We extended this concept by demonstrating that the severity of gastric damage caused by a common daily dose of 325 mg ASA is reduced when administered together with NCX-4016. Similar to ASA, administration of NCX-4016 resulted in near complete inhibition of gastric PGE2, suggesting that NO has the ability to compensate for the reduced generation of gastric prostanoids (31,32). This observation might be of relevance when multiple antiplatelet drugs, with potential gastric toxicity, are co-administered to prevent restenosis to patients that underwent coronary or carotid angioplasty. Low doses of ASA (i.e., 100 mg/day or less) and clopidogrel are currently used to minimize gastric toxicity, and are similarly effective and tolerable (33). At present, no studies have yet compared NCX-4016 and clopidogrel in terms of efficacy and gastric tolerability. However, the data from the present study permit the speculation that NCX-4016 may be at least as tolerable as clopidogrel.
Finally, NCX-4016 does not induce in humans any significant change in blood pressure and heart rate, in agreement with previous observations and with the notion that NO is slowly released from the molecule (P. del Soldato, personal communication, June 2003).
In conclusion, the results of the present study indicate that NCX-4016 exhibits a broader spectrum of activities than ASA and that NCX-4016 induces less gastric damage also when administered in combination with the former. This study is the first demonstration that a NO-donor and COX inhibitor downregulates TF as well as IL-6 and MCP-1 in humans. These beneficial effects of NCX-4016, combined with the inhibition of TXA2, support the notion that NCX-4016 exerts multiple regulatory activities at the monocyte/platelet interface and may be of relevance in the clinical setting of atherothrombotic diseases and restenosis after angioplasty, where TF expression and platelet-monocyte interactions together with increased generation of TXA2are observed (1,6). Larger multicenter clinical trials are warranted in order to establish the clinical efficacy and safety of NCX4016.
The authors thank Erminio Bonizzoni, PhD, for his advice in the statistical analysis and Alberto Fransioli, MD, for his help in performing endoscopies.
Drs. Fiorucci and Meneguzzi received a grant from NicOx. Dr. del Soldato works for NicOx as an Executive Vice President, Science and Technology. This study was supported by NicOx S.A., Sophia Antipolis, France.
- Abbreviations and Acronyms
- aspirin/acetylsalicylic acid
- nitric oxide
- proteinase activated receptor
- tissue factor
- thrombin receptor activator protein
- Received January 30, 2004.
- Revision received February 27, 2004.
- Accepted March 2, 2004.
- American College of Cardiology Foundation
- Riewald M.,
- Ruf W.
- Corseaux D.,
- Le Tourneau T.,
- Six I.,
- et al.
- Toschi V.,
- Gallo R.,
- Lettino M.,
- et al.
- Cunningham M.A.,
- Romas P.,
- Hutchinson P.,
- Holdsworth S.R.,
- Tipping P.G.
- Fiorucci S.,
- Santucci L.,
- Cirino G.,
- et al.
- Fiorucci S.,
- Mencarelli A.,
- Meneguzzi A.,
- et al.
- Catella-Lawson F.,
- Reilly M.P.,
- Kapoor S.C.,
- et al.
- Eikelboom J.W.,
- Hirsh J.,
- Weitz J.I.,
- Johnston M.,
- Yi Q.,
- Yusuf S.
- Cipollone F.,
- Patrignani P.,
- Greco A.,
- et al.
- Fiorucci S.,
- Santucci L.,
- Wallace J.L.,
- et al.
- Barrios-Rodiles M.,
- Tiraloche G.,
- Chadee K.
- Yang Y.,
- Loscalzo J.
- Lindmark E.,
- Tenno T.,
- Siegbahn A.
- Schober A.,
- Manka D.,
- von Hundelshausen P.,
- et al.
- Mori E.,
- Komori K.,
- Yamaoka T.,
- et al.
- Schecter A.D.,
- Rollins B.J.,
- Zhang Y.J.,
- et al.
- Kitamoto S.,
- Egashira K.,
- Kataoka C.,
- et al.