Author + information
- Received December 21, 2009
- Revision received February 8, 2010
- Accepted February 9, 2010
- Published online June 15, 2010.
- Julinda Mehilli, MD*,
- Robert A. Byrne, MB*,
- Klaus Tiroch, MD*,
- Susanne Pinieck*,
- Stefanie Schulz, MD*,
- Sebastian Kufner, MD*,
- Steffen Massberg, MD*,
- Karl-Ludwig Laugwitz, MD†,
- Albert Schömig, MD†,
- Adnan Kastrati, MD*,* (, )
- ISAR-DESIRE 2 Investigators
- ↵*Reprint requests and correspondence:
Dr. Adnan Kastrati, Deutsches Herzzentrum, Lazarettstrasse 36, Munich 80636, Germany
Objectives For patients with sirolimus-eluting stent (SES) restenosis requiring reintervention, we compared a strategy of repeat SES (Cypher, Cordis, Miami Lakes, Florida) implantation with paclitaxel-eluting stent (PES) (Taxus, Boston Scientific, Natick, Massachusetts) implantation.
Background Despite their high anti-restenotic efficacy, the widespread utilization of SES therapy has led to a significant absolute number of patients presenting with SES treatment failure. The optimal treatment strategy for such patients remains unclear.
Methods The ISAR-DESIRE 2 (Intracoronary Stenting and Angiographic Results: Drug Eluting Stents for In-Stent Restenosis 2) study was a randomized, open-label, active-controlled trial conducted among 450 patients with clinically significant in-SES restenosis at 2 centers in Munich, Germany. After pre-treatment with 600 mg clopidogrel, all patients were randomly assigned to either SES or PES implantation. The primary end point was late lumen loss, based on in-stent analysis, at 6- to 8-month follow-up angiography. Secondary end points were binary angiographic restenosis (diameter stenosis >50%) at 6- to 8-month follow-up, target lesion revascularization, the composite of death or myocardial infarction, and definite stent thrombosis at 12 months.
Results Regarding anti-restenotic efficacy, there were no differences between SES and PES in late loss (0.40 ± 0.65 mm vs. 0.38 ± 0.59 mm; p = 0.85), binary restenosis (19.6% vs. 20.6%; p = 0.69), or target lesion revascularization (16.6% vs. 14.6%; p = 0.52). In terms of safety outcomes, the rates of death/myocardial infarction (6.1% vs. 5.8%; p = 0.86) and stent thrombosis (0.4% vs. 0.4%; p > 0.99) were also similar.
Conclusions In cases of SES restenosis, treatment with either repeat SES or switch to PES was associated with a comparable degree of efficacy and safety. Drug resistance at an individual patient level may play a contributory role to the somewhat higher than expected late loss observed with the SES in the current study. (Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for In-Stent Restenosis 2 [ISAR-DESIRE 2]; NCT00598715)
Drug-eluting stent (DES) therapy has undoubtedly represented a significant success in the battle against coronary artery restenosis (1). The widespread adoption of this technology into routine clinical practice, however, has led to significant absolute numbers of patients presenting with DES treatment failure. In the U.S. alone, it is estimated that as many as 200,000 cases of DES restenosis may occur every year (2).
For patients with DES restenosis, the optimal management strategy remains unknown. The predominantly focal pattern of restenosis (3–5) and the lessons learned from the management of bare metal stent restenosis—where DES therapy proved superior to balloon angioplasty and vascular brachytherapy (6–9)—makes treatment with repeat DES an attractive option. However, it remains to be determined whether implantation of DES eluting the same type of drug—a so-called “homo-DES” approach—is preferable to a switch to a DES eluting a different class of drug—a “hetero-DES” strategy.
The specific setting of sirolimus-eluting stent (SES) restenosis is a case in point, and its management a matter of clinical equipoise. On the one hand, it is conceivable that patients with SES treatment failure may possess specific factors (such as hyporesponsiveness to sirolimus) that would favor a switch to treatment with a paclitaxel-eluting stent (PES). On the other hand, the superior neointimal suppression seen with first-generation SES as compared with PES may be particularly apposite in the setting of a high risk of recurrent restenosis. To investigate this issue, we randomly assigned patients with SES restenosis to repeat SES implantation (Cypher, Cordis, Miami Lakes, Florida) or to switch to PES therapy (Taxus, Boston Scientific, Natick, Massachusetts).
Study population, randomization, and intervention protocol
Between October 2007 and January 2009, patients >18 years of age with ischemic symptoms or evidence of myocardial ischemia (inducible or spontaneous) in the presence of a restenosis ≥50% located in the native vessel segment treated with SES were considered eligible, provided that written, informed consent by the patient or a legally authorized representative for participation in the study was obtained. Patients with restenosis occurring in either the sirolimus-eluting Cypher stent (Cordis) or the sirolimus-eluting ISAR (individualizable stent to abrogate restenosis) stent (based on Yukon [Translumina, Hechingen, Germany] backbone) (10) were considered eligible for participation in the study. Patients with a target lesion located in the left main stem, acute myocardial infarction within the preceding 48 h, cardiogenic shock, malignancies, or other comorbid conditions with life expectancy <12 months or that may result in protocol noncompliance, known allergy to the study medications (sirolimus, paclitaxel), or pregnancy (present, suspected, or planned) were considered ineligible for the study. The study was conducted in accordance with the provisions of the Declaration of Helsinki and with the International Conference on Harmonization Good Clinical Practices. The trial protocol was approved by the institutional ethics committee responsible for the participating centers, Deutsches Herzzentrum and 1. Medizinische Klinik, Klinikum Rechts der Isar, both in Munich, Germany.
In each participating center, allocation to treatment was made by means of sealed, opaque envelopes containing a computer-generated sequence; randomization was performed immediately after the decision to proceed with percutaneous coronary intervention (PCI). Patients who met all of the inclusion criteria and none of the exclusion criteria were randomized in the order that they qualified. Randomization was stratified according to participating center. Patient allocation to each of the 2 treatment groups was in equal proportions. Both treatment groups were studied concurrently. Time zero was defined as the time of randomization, and patients were considered enrolled in the study at this time point. The same randomly assigned stent had to be implanted in all lesions in patients who required stenting in multiple lesions, and the use of >1 stent per lesion was also allowed. Patients were assigned to receive either the sirolimus-eluting Cypher (Cordis) stent or paclitaxel-eluting Taxus stent (Boston Scientific).
An oral loading dose of 600 mg clopidogrel was administered to all patients before the intervention, regardless of whether the patient was receiving clopidogrel before admission. During the procedure, patients were given intravenous aspirin, heparin, or bivalirudin; glycoprotein IIb/IIIa inhibitor usage was at the discretion of the operators. Procedural success was defined as residual stenosis <30% and Thrombolysis In Myocardial Infarction flow grade 3 achieved in the treated vessel using the assigned stent. After the intervention, all patients, irrespective of treatment allocation, were prescribed 200 mg/day aspirin indefinitely, clopidogrel 150 mg for the first 3 days (or until discharge) followed by 75 mg/day for at least 6 months, and other cardiac medications according to the judgment of the patient's physician (e.g., beta-blockers, angiotensin-converting enzyme inhibitors, statins). After enrollment, patients remained in the hospital for at least 48 h. Blood samples were drawn every 8 h for the first 24 h after randomization and daily afterward for the determination of cardiac markers (creatine kinase, creatine kinase-myocardial band, troponin T). Daily recording of the electrocardiogram was also performed until discharge. All patients were evaluated at 1 and 12 months by phone or office visit. Repeat coronary angiography was scheduled for all patients at 6 to 8 months.
Data management, end points, and definitions
Relevant data were collected and entered into a computer database by specialized personnel of the clinical data management center. All events were adjudicated and classified by an event adjudication committee blinded to the treatment groups. Baseline, post-procedural, and follow-up coronary angiograms were digitally recorded and assessed off line in the quantitative angiography (QCA) core laboratory (ISARESEARCH Center, Munich, Germany) with an automated edge-detection system (CMS version 7.1, Medis Medical Imaging Systems, Ridgefield, Connecticut) by 2 independent experienced operators unaware of the treatment allocation. Measurements were performed on cineangiograms recorded after the intracoronary administration of nitroglycerine. Baseline QCA measurements were performed using the single worst view projection for the index lesion; the same view projection was used for the measurements after stent implantation. In the follow-up angiogram, the QCA measurements were performed using the single worst view projection at that time point. The contrast-filled nontapered catheter tip was used for calibration. Quantitative analysis was performed on both the “in-stent” and “in-segment” area (including the stented segment, as well as both 5-mm margins proximal and distal to the stent). Restenosis morphology was adjudicated according to criteria modified from Mehran et al. (11) Restenosis was defined as diameter stenosis ≥50% in the in-segment area (including stent area as well as 5-mm margins proximal and distal to the stent). At baseline angiogram of the index lesion, the stent was defined as the previously implanted SES. At follow-up angiogram of the recurrent restenotic lesion, the stent was defined as the new implanted stent in the setting of the present study (SES or PES).
The primary end point of the study was in-stent late lumen loss at follow-up angiography (defined as the difference between the minimal luminal diameter at the end of the procedure and the minimal luminal diameter at follow-up angiography). Secondary end points were in-segment binary angiographic restenosis, defined as diameter stenosis ≥50% in the in-segment area (including the stent area as well as 5-mm margins proximal and distal to the stent) at follow-up angiography; the need for target lesion revascularization (TLR), defined as any revascularization procedure involving the target lesion due to luminal renarrowing in the presence of symptoms or objective signs of ischemia at 1-year follow-up; the combined incidence of death or myocardial infarction; and the incidence of definite stent thrombosis. A detailed definition of myocardial infarction has been previously described (12). Stent thrombosis was classified according to Academic Research Consortium criteria (13).
The objective of the study was to assess the superiority of PES compared to SES for prevention of recurrent restenosis in patients who had undergone a prior SES implantation. The null hypothesis regarding the primary end point was that there was no difference between the SES and the PES. Sample size calculation was based on the method for 2 independent means using the following assumptions: in-stent late luminal of 0.60 mm after repeated SES implantation, in-stent late lumen loss of 0.40 mm after PES implantation, a common standard deviation of 0.60 mm, 2-sided alpha-level of 0.05, and power of 90%. Accordingly, we estimated that 190 patients with angiographic follow-up data per group were needed. To account for possible losses to follow-up, a total of 450 patients was enrolled. The analysis of primary and secondary end points was planned to be performed on an intention-to-treat basis. Pre-specified subsets of interest were old and young patients, men and women, diabetic and nondiabetic patients, and small and large vessels. Continuous data are presented as mean (SD) or median (25th to 75th percentiles). Categorical data are presented as counts or proportions (%). Differences between groups were checked for significance using Student ttest for continuous data and chi-square test (or Fisher's exact test where the expected cell value was <5) for categorical variables. Survival was assessed using the methods of Kaplan-Meier and compared using the log-rank test. Statistical software S-PLUS version 4.5 (S-PLUS, Insightful Corp., Seattle, Washington) was used for analysis. Sample size calculation was performed using nQuery Advisor (Statistical Solutions, Cork, Ireland).
Patient and lesion characteristics
A total of 450 patients was enrolled in the ISAR-DESIRE 2 study and randomly assigned to receive either SES (n = 225) or PES (n = 225). They represent 90.2% of the 499 eligible patients treated for restenosis. Of the 49 patients not enrolled in the study, 9 patients presented with acute myocardial infarction, 30 with in-stent restenosis of the left main area, and 11 patients refused study participation. Baseline characteristics of the study patients are shown in Table 1,and they were well matched between the 2 treatment groups. Diabetes mellitus was present in 162 (36.0%) patients. Eighty-three (18.4%) patients presented with acute coronary syndrome.
Angiographic and procedural characteristics are displayed in Table 2.There were no significant differences between the 2 treatment arms. In all, 243 lesions were treated in the SES arm versus 240 in the PES arm (p = 0.61).
Restenosis stent type and morphology
In the SES group, 94 (38.7%) patients had received a Cypher SES at the time of the initial procedure, with 149 (61.3%) receiving an ISAR SES. Of patients allocated to PES treatment, 92 (38.3%) had received a Cypher stent, with 148 (61.7%) treated with an ISAR SES (p = 0.94). Overall, there were no significant differences between the groups with regard to restenosis morphology at presentation. The majority presented with a focal pattern of restenosis (SES n = 158 [65.0%] vs. PES n = 146 [60.8%]; p = 0.34) (details in Table 2). Procedural success was achieved in 242 (99.6%) lesions with SES versus 249 (100.0%) with PES (p = 0.32).
Angiographic outcomes and recurrent restenosis morphology
Angiographic follow-up data were available for 190 (84.8%) patients and 191 (84.9%) patients treated with SES and PES, respectively (p = 0.98). Median time to surveillance angiography was 195 days (168 to 208 days) and 198 days (164 to 217 days), respectively (p = 0.22). Angiographic results and the patterns of recurrent restenosis are displayed in Table 3.
In terms of the primary angiographic end point, there was no significant difference between the 2 groups. Mean late loss was 0.40 ± 0.65 mm in the SES group against 0.38 ± 0.58 mm in the PES group (p = 0.85) (Fig. 1).Similarly, there were no significant differences between the groups in terms of binary restenosis (SES n = 39 [19.0%] lesions versus PES n = 42 [20.6%] lesions; p = 0.69) (Fig. 2).
Regarding the analysis of the primary end point in pre-specified subgroups, lack of difference between the 2 stent types was observed in all pre-specified subgroups defined by age, sex, diabetes, and vessel size (p value for interaction >0.15 in all cases). Neither were there any differences between the 2 groups when patients were analyzed according to the stent type implanted at the initial intervention (p value for interaction = 0.49). Among patients originally treated with a Cypher SES, mean late loss with a second SES was 0.40 ± 0.61 mm versus 0.43 ± 0.64 mm with PES (p = 0.73). Among patients initially treated with a polymer-free SES, late loss with Cypher SES was 0.40 ± 0.68 mm versus 0.34 ± 0.56 mm with PES (p = 0.49).
At 12 months, there were no significant differences between the 2 groups in terms of clinical results (Table 4).TLR was required in 35 (16.6%) patients treated with SES as compared with 30 (14.6%) patients treated with PES (p = 0.52) (Fig. 2). Two (0.8%) patients treated with SES were in need of bypass surgery as compared with 1 (0.4%) patient treated with PES (p = 0.56). Target vessel revascularization was performed in 50 (23.7%) patients treated with Cypher versus 47 (22.8%) patients treated with Taxus (p = 0.82).
Regarding safety end points, there were no differences in the composite of death or myocardial infarction (SES 13 [6.1%] vs. PES 12 [5.8%]; p = 0.86) or the rate of definite stent thrombosis (SES 1 [0.4%] vs. PES 1 [0.4%]; p = 0.67). Both stent thrombosis events occurred after 30 days.
In terms of overall major adverse cardiovascular events, there was no difference in the composite of death, myocardial infarction, or target lesion revascularization (SES n = 44 [20.4%] vs. PES n = 41 [19.6%]; p = 0.71) (Fig. 3).
In relation to antiplatelet therapy, 61.2% of patients treated with SES and 64.9% of patients treated with PES were still receiving dual antiplatelet therapy 1 year after intervention (p = 0.49). For patients who stopped dual antiplatelet therapy, median time to clopidogrel discontinuation was 223 days (range 188 to 340 days) for patients treated with SES and 239 days (range 200 to 344 days) for patients treated with PES (p = 0.68).
The principal findings of the ISAR-DESIRE 2 study are that in cases of DES restenosis occurring within a SES: 1) the implantation of a second DES is feasible and safe; 2) a strategy of either repeat SES implantation or switch to a PES is associated with comparable anti-restenotic efficacy; and 3) the neointimal inhibition observed with repeat SES implantation is somewhat lower than expected, indicative perhaps of the existence of some degree of drug hyporesponsiveness at an individual patient level.
The widespread adoption of DES therapy coupled with an overall increase in the number of PCI procedures has generated significant absolute numbers of patients presenting with DES restenosis (2). Clinically, the treatment of these DES restenotic lesions seems to be associated with slightly higher rates of recurrent restenosis than that observed after treatment of bare-metal stent restenosis. For example, the use of DES to treat bare-metal stent restenosis has been associated with repeat TLR rates at or just below 10% in most studies (6–8,14). Conversely, the use of DES to treat DES restenosis carries with it a TLR rate in the range of 10% to 20% (3–5,14–17). Pathological differences in plaque morphology as compared with bare-metal stent restenosis may also exist—for example, obstructive DES restenotic plaques may be characterized by a proliferative expansion of cell-depleted amorphous extracellular matrix and a possible excess of in-stent atherosclerosis (18).
Catheter-based treatment of bare-metal stent in-stent restenosis was most effectively accomplished by DES implantation. In the setting of randomized controlled trials, this proved superior to both plain balloon angioplasty (6,9) and vascular brachytherapy (7,8). Nevertheless, for patients with DES restenosis, the optimal management approach remains to be defined.
As the predominant pattern of restenosis within DES is focal—exemplified by approximately two-thirds of cases in the current study—catheter-based intervention would appear to be the most attractive initial treatment option for this disease. In this respect, the optimal balance between maximal acute gain and minimal late loss is likely to remain with repeat DES implantation. The current study adds to earlier registry experience on this strategy and indicates that repeat DES implantation is a feasible and safe approach. Although concern has always existed regarding the advisability of the implantation of multiple stent layers in the same arterial segment, no evidence of any adverse safety signal was observed out to 12 months, with a rate of definite stent thrombosis of 0.4% for the study population as a whole. In terms of clinical efficacy, the overall rate of TLR at 12 months (approximately 15%) is remarkably consistent with that of earlier nonrandomized studies (3–5,15–17) and was not different between the SES and PES treatment groups. This latter observation was mirrored in the comparative magnitude of late loss (SES 0.40 ± 0.65 mm vs. PES 0.38 ± 0.59 mm) between the 2 stent platforms and in keeping with the null hypothesis of the study (H0not rejected, p = 0.85).
These findings have at least 2 implications. First of all, from a practical point of view, when faced with a patient with SES restenosis requiring repeat intervention, the clinician may chose between a strategy of either repeat SES implantation or switch to PES therapy, with an expectation of similar clinical outcome from either approach. Second, from a theoretical standpoint, the findings are unusual in certain respects as the Cypher stent has tended to outperform the Taxus stent in higher-risk patient subsets (6,19–21). The equivalent efficacy of both stents is most likely related to a relative underperformance of the Cypher SES, which might be related to sirolimus hyporesponsiveness at an individual patient level in the population under study. As evidence of this, the angiographic anti-restenotic efficacy of the SES is somewhat poorer than would be expected from the results of previous studies that enrolled enriched patient populations. In the ISAR-DESIRE study, the Cypher SES showed a mean in-stent late luminal loss of 0.21 ± 0.59 mm in patients with bare-metal stent restenosis (6). In diabetic patients studied in the ISAR-DIABETES trial, the same stent was associated with a late loss of 0.19 ± 0.44 mm (19); and in the setting of small vessels (20), the value was 0.23 ± 0.55 mm. Notwithstanding the difficulties encountered in historical comparison, the mean late loss observed with the SES in the current study—0.40 ± 0.65 mm—is an outlier and supports the existence of relative resistance to sirolimus in patients presenting with SES restenosis.
Resistance to sirolimus is well described in the oncology literature. The drug exerts its antiproliferative effects by complexing with FKBP12 and inhibiting the function of the mammalian target of rapamycin (mTOR)—an evolutionarily conserved protein kinase with a key role in cell growth, proliferation, and survival. Resistance may be conferred by mutations either in the FKBP12 (preventing initial complexing) or in the FRB domain of mTOR (which inhibits binding of the drug-protein complex to the mTOR receptor) (22). Furthermore, downstream resistance linked to defective regulation of 3 key mTOR signaling pathways—4E-BP1, S6K1, and p27kip—has also been described (22). The existence of a similar hyporesponsiveness phenomenon after local drug delivery to the coronary vessel wall would not be entirely unexpected.
Some important issues in the management of DES restenosis remain to be addressed. First, whether these results are applicable to the treatment of restenosis occurring within newer DES is open to discussion. This question could be definitively answered by specifically designed randomized trials. Second, if ability to identify drug resistance in particular patients develops, this might conceivably be used to guide therapy. Third, although results with drug-eluting balloon therapy show promise in cases of bare-metal stent restenosis (23), the place of this therapy in cases of DES restenosis awaits definition and will be the subject of future investigation (the ISAR-DESIRE-3 study; clinicaltrials.gov identifier NCT00987324).
As with any study utilizing an angiographic end point, conclusions are based on incomplete data observations. In this respect, it is notable that the overall angiographic follow-up rate (almost 85%) was above the threshold at which such end points are considered to be robust surrogates of device efficacy (24). In addition, the optimal time point for adjudication of DES anti-restenotic efficacy remains to be defined, as evidence suggests that ongoing delayed late luminal loss beyond the usual 6- to 8-month time window is a feature of first-generation DES therapy (25). Furthermore, the influence of angiographic follow-up on the absolute rate of TLR should be considered. This may increase the incidence of TLR in a manner that may not reflect routine clinical practice, although the relative magnitude of an observed treatment effect may be expected to be real (26). Additionally, 60% of the patients were treated because of restenosis in an ISAR-SES, a stent that is not widely available. Although the late loss for Taxus was numerically slightly lower among patients in whom the initial stent implanted was an ISAR-SES, as distinct from a Cypher SES, there was no statistically significant interaction between treatment assignment and type of SES implanted at the initial intervention. Finally, regarding safety outcomes, this study was not powered to detect a difference in rarely occurring clinical events such as stent thrombosis.
The findings of the ISAR-DESIRE 2 study demonstrate that, in cases of SES restenosis, both SES and PES are associated with a comparable degree of anti-restenotic efficacy and clinical safety. The somewhat higher than expected late loss with the SES in the current study suggests that drug resistance at an individual patient level plays a contributory role in SES restenosis.
Steering Committee: A. Schömig (Chairman), A. Kastrati (Principal Investigator), and J. Mehilli. Participating Centers:Deutsches Herzzentrum, Technische Universität, Munich, Germany, and 1. Medizinische Klinik, Klinikum Rechts der Isar, Technische Universität, Munich, Germany. Clinical Events Adjudication Committee:D. Hall (Chairman, deceased), G. Ndrepepa, and L. Goedel-Meinen. ISARESEARCH Center Data Coordination:J. Mehilli (Director), K. A. Birkmeier, M. Dirlewanger, B. Gliessl, H. Holle, S. Kufner, K. Hösl, N. Rifatov, F. Maimer-Rodrigues, N. Sargon, and S. Schulz. ISARESEARCH Center Angiographic Core Laboratory:A. Bergbauer, O. Bruskina, R. A. Byrne, S. Hurt, R. Iijima, S. Pinieck, and S. Ranftl.
Dr. Mehilli has received lecture fees from Cordis. Dr. Byrne was supported by a Research Fellowship in Atherothrombosis from the European Society of Cardiology. Dr. Kastrati has received lecture fees from Cordis and Medtronic.
The study design, analysis, and funding were performed by Deutsches Herzzentrum, Munich, and was industry independent.
An abstract of this work was presented as a Late Breaking Clinical Trial at Transcatheter Cardiovascular Therapeutics 2009 in San Francisco, California.
- Abbreviations and Acronyms
- drug-eluting stent(s)
- individualizable stent to abrogate restenosis
- percutaneous coronary intervention
- paclitaxel-eluting stent(s)
- quantitative angiography
- sirolimus-eluting stent(s)
- target lesion revascularization
- Received December 21, 2009.
- Revision received February 8, 2010.
- Accepted February 9, 2010.
- American College of Cardiology Foundation
- Lloyd-Jones D.,
- Adams R.,
- Carnethon M.,
- et al.
- Alfonso F.,
- Perez-Vizcayno M.J.,
- Hernandez R.,
- et al.
- Mehilli J.,
- Kastrati A.,
- Wessely R.,
- et al.
- Mehran R.,
- Dangas G.,
- Abizaid A.S.,
- et al.
- Byrne R.A.,
- Kastrati A.,
- Kufner S.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Mishkel G.J.,
- Moore A.L.,
- Markwell S.,
- Shelton M.C.,
- Shelton M.E.
- Nakazawa G.,
- Vorpahl M.,
- Finn A.V.,
- Narula J.,
- Virmani R.
- Mehilli J.,
- Dibra A.,
- Kastrati A.,
- Pache J.,
- Dirschinger J.,
- Schömig A.
- Kuntz R.E.,
- Keaney K.M.,
- Senerchia C.,
- Baim D.S.
- Byrne R.A.,
- Iijima R.,
- Mehilli J.,
- et al.
- Pinto D.S.,
- Stone G.W.,
- Ellis S.G.,
- et al.