Influence of a Nonionic, Iso-Osmolar Contrast Medium (Iodixanol) Versus an Ionic, Low-Osmolar Contrast Medium (Ioxaglate) on Major Adverse Cardiac Events in Patients Undergoing Percutaneous Transluminal Coronary Angioplasty
A Multicenter, Randomized, Double-Blind Study
Background—The potential merits and disadvantages of the use of ionic or nonionic contrast media in patients undergoing percutaneous transluminal coronary angioplasty (PTCA) have been the subjects of controversy. The present study was designed to evaluate the possible influence of both types of contrast media on major adverse cardiac events (MACE) in patients undergoing PTCA.
Methods and Results—In a randomized, parallel-group, double-blind study, 1411 patients received either iodixanol (a nonionic, iso-osmolar contrast medium) or ioxaglate (an ionic, low-osmolar contrast medium) during PTCA. A standardized anticoagulation regimen was followed. Patients were monitored in the hospital for 2 days and followed-up at 1 month. The primary end point, a composite of MACE (death, stroke, myocardial infarction, coronary artery bypass grafting, and re-PTCA) after 2 days, occurred in 4.3% of the total population, with no statistically significant difference between groups (iodixanol, 4.7%; ioxaglate, 3.9%; P=0.45). Further, between 2-day and 1-month follow-ups, no significant difference (P=0.27) existed between the groups in the rates of MACE. Hypersensitivity reactions (P=0.007) and adverse drug reactions (P=0.002) were significantly less frequent in the iodixanol group. The only significant predicting factors for the occurrence of MACE were dissection/abrupt closure and country.
Conclusions—No significant differences were observed between the iodixanol and ioxaglate groups with regard to MACE, although hypersensitivity and adverse drug reactions were significantly less frequent in patients who received iodixanol.
The potential merits and disadvantages of ionic versus nonionic contrast media during percutaneous transluminal coronary angioplasty (PTCA) and the influences of these media on clinical outcome are controversial.1 2 3 4 5 6 7 8 Thrombosis and vessel dissection may influence the clinical outcome after coronary angioplasty, a procedure that invariably damages the endothelium and produces varying degrees of medial injury. It has been suggested that ionic contrast media may be advantageous in this setting because they can act as anticoagulants and inhibitors of platelet aggregation, whereas nonionic contrast media do not have these effects.1 2 3 4 5 6 7 8 However, results from preclinical studies are equivocal.9 10 11 12 13 14 15 16 17 18 19 20 In addition, clinical data are limited, and no study has shown an increase in major adverse cardiac events (MACE) after the administration of nonionic compared with ionic contrast media. Therefore, we designed a large-scale clinical study to evaluate the possible effects of an ionic (ioxaglate) and a nonionic (iodixanol) contrast medium on clinical outcome and MACE in patients undergoing PTCA. The primary end point was a composite of MACE, defined as death, stroke, myocardial infarction (MI), and the need for coronary artery bypass grafting (CABG) or repeated PTCA (re-PTCA). The study population was composed of patients who had stable or unstable angina or silent ischemia.
A multicenter, parallel-group, randomized, double-blind study was conducted in 32 centers in 5 European countries (Belgium, France, Spain, Sweden, and The Netherlands). It was designed to compare the effects of the nonionic, iso-osmolar contrast medium iodixanol 320 mg per I/mL (Visipaque, Nycomed Imaging AS) and the ionic, low-osmolar contrast medium ioxaglate 320 mg I/mL (Hexabrix, Guerbet SA) in patients undergoing PTCA. Intracoronary stenting was allowed, and it was classified as planned or unplanned. The aim of the study was to assess the influence of the choice of contrast medium and of other clinical and procedural factors (eg, clinical presentation, age, sex, and stenting condition) on the occurrence of MACE after PTCA. Anticoagulation was standardized and monitored during the procedure by repeated measurements of the activated clotting time (ACT). The patients were randomized to receive 1 of the 2 contrast media, which were placed in identical vials to achieve double blinding. The local ethics committees and the national regulatory authorities of each institution and country approved the protocol. Informed written consent was obtained from each patient before enrollment in the trial.
Between November 21, 1996, and September 22, 1997, 1541 patients ≥18 years of age who had stable (Canadian Cardiovascular Society functional classification) or unstable angina (Braunwald’s classification system) or silent ischemia and who were undergoing PTCA were enrolled in the study. Exclusion criteria were recent (<7 days) acute MI, unprotected left main stenosis, or the need for oral anticoagulation with antivitamin K. Patients were also excluded if they had contraindications to heparin, aspirin, or ticlopidine therapy or to iodinated contrast media. Patients who had received iodinated contrast media within the 18 hours before PTCA were not included to minimize the possible influence of nonstudy contrast media on the results. The use of rotablator, directional atherectomy, or pretreatment with glycoprotein IIb/IIIa receptor inhibitors was not allowed. Left ventricular ejection fraction was required to be >35%. In a few centers, recruitment of patients was performed before diagnostic catheterization, and trial contrast medium was then used for that purpose to ensure identical contrast media throughout the procedures. Among the 1541 patients enrolled, 81 did not have PTCA, 27 had major protocol violations (the majority due to an ejection fraction <35% or to the presence of a recent MI), 26 were excluded because they received a mixture of different contrast media, 3 were excluded due to loss of their case report form or contrast medium vials, and 7 were excluded for >1 of the above-mentioned reasons. Of the 1411 patients who constituted the per-protocol population, 697 received iodixanol and 714 received ioxaglate.
All patients received heparin, and all but 4 received an antiplatelet agent (≥100 mg of aspirin and/or ticlopidine). An intravenous bolus of heparin (10 000 IU) was administered at the start of the procedure. The ACT was measured 5 minutes after the first injection of heparin, and further heparin was given if necessary to maintain the ACT above 250 or 300 s (depending on the apparatus used). If PTCA lasted >1 hour, ACT was recorded again, and additional heparin was given when required. ACT was measured again at the end of the procedure. Ticlopidine was given to patients after stent implantation or if otherwise indicated (eg, allergy to aspirin). Overall, 68.6% of patients received ticlopidine either before, as routine medication, or before/during/after PTCA. All other medications and procedures were at the discretion of the investigator.
Procedural and safety outcomes were recorded for all patients during the hospital stay (2-day follow-up). The primary end point was a composite of MACE during the 2-day follow-up period (2-day MACE); it was defined as death, stroke, Q-wave or non-Q-wave MI (NQWMI), and CABG and/or emergency re-PTCA at the target lesion or in the target vessel. The definition of a NQWMI was an increase in serum creatine kinase (CK) to twice the upper normal reference range limit or more, coupled with an increase in the CK-isoenzyme MB fraction. An independent end point classification committee evaluated CK for all patients to determine NQWMI cases. This independent end point classification committee, which consisted of 3 cardiologists, also evaluated, in a blinded fashion, all reported MACE-2 day cases. The committee’s decision was based on information from the case report form, ECGs, vital signs, laboratory test results (pre- and post-PTCA), and films/videos from the PTCA procedure. In addition to the in-hospital follow-up, the patients were contacted by telephone (by the investigator or a study nurse) 1 month after discharge from the hospital to ascertain if they had been rehospitalized due to MACE. The end point classification committee did not evaluate occurrence of MACE between the 2-day and 1-month follow-ups.
To assess the influence of contrast media and other factors on the occurrence of 2-day MACE, the factors were individually evaluated through odds ratio (OR) estimates and their associated 95% confidence intervals (single-factor analyses). Testing OR deviations from 1 was done through the 2-tailed likelihood χ2 test; P≤0.05 was considered statistically significant. In addition, an exploratory multiple logistic regression analysis was performed based on the results of the single-factor analysis to identify the predictors for 2-day MACE; it took into account possible interactions between factors. Other adverse events, including hypersensitivity reactions, were compared between contrast media groups at 2 days and at 1 month using the χ2 test.
The comparability of the 2 contrast media groups with regard to baseline characteristics was assessed through the dependency of the primary end point on each baseline characteristic. For each baseline characteristic, the χ2 P value for testing independence between the baseline characteristic and primary end point was calculated per contrast medium group; P<0.05 indicates dependency. In cases where dependency is found, a large difference in the baseline characteristics between the contrast media groups might then have unwanted effects on the comparability of the groups.
The baseline clinical characteristics of the study population are presented in Table 1⇓ and were similar in both contrast media groups. The indications for PTCA and the location of the treated lesions did not differ between the groups. Baseline angiographic characteristics are given in Table 2⇓. No significant relationships existed between the baseline characteristics listed in Tables 1⇓ and 2⇓ and 2-day MACE. Specifically, for presence of diabetes and family history of coronary artery disease, the 2 factors for which frequencies between the contrast medium groups deviated most, the probability values for testing of independence were all >0.74. Hence, these random deviations had no influence on the rate of 2-day MACE in the contrast media groups.
The majority of the patients underwent single-vessel PTCA (73.0% versus 71.1%, iodixanol group versus ioxaglate group). A total of 23% of the patients (22.2% in the iodixanol group versus 23.4% in the ioxaglate group) had 2-vessel PTCA, whereas the remaining patients underwent 3-vessel PTCA. Glycoprotein IIb/IIIa receptor inhibitors were given during or after the PTCA procedure in 23 patients (12 iodixanol and 11 ioxaglate) who had abrupt closure, stent occlusion, or coronary dissection.
The percent diameter stenosis at baseline and after PTCA are reported in Table 2⇑. No difference in percent residual stenosis existed between the 2 groups immediately after PTCA. The proportion of angiographically identifiable thrombus (visual assessment) before and after PTCA was similar in both contrast media groups (all lesions included) (Table 2⇑). The rates of abrupt closure (all lesions) occurring during the procedure did not differ significantly between groups (P=0.39). Procedure-related dissections were reported somewhat less frequently in patients allocated to the iodixanol group compared with the ioxaglate group, but the difference was not statistically significant (P=0.18) (Table 2⇑). Implantation of intracoronary stents was similar in both groups: 59.8% in the iodixanol group and 60.9% in the ioxaglate group. The implantation rate varied from 36.0% in the country with the lowest stent implantation rate (France) to 76.9% in the country with the highest rate (Belgium).
Primary End Point
On the basis of the intention-to-treat population (1538 patients, 765 in iodixanol group and 773 in ioxaglate group), which consisted of all enrolled patients but the 3 who had a lost case report form or contrast medium vials, the primary end point (2-day MACE) occurred in 4.2%, with no statistically significant differences between the iodixanol and the ioxaglate groups (4.6% versus 3.8%, respectively; P=0.42). Among the MACE, NQWMI was the most frequent complication (2.7%). The analysis of the per-protocol population (1411 patients) showed similar results (Table 3⇓). No statistically significant differences existed between the iodixanol and the ioxaglate groups (4.7% versus 3.9%; P=0.45). Among the 2-day MACE, NQWMI was the most frequent complication (2.9%). The majority (76%) of these patients had a >4-fold increase in CK values. Two cardiac deaths occurred during the 2-day follow-up period, both in the ioxaglate group. The 2-day MACE rate varied among the centers, from 0% to 18.2% among those completing their inclusion. One center enrolled only 7 patients and reported 2 patients with 2-day MACE. Patients who underwent stent implantation had a 4.6% chance of developing 2-day MACE; the chance was 3.3% when stenting was planned and 5.3% when it was unplanned. The rate for nonstented patients was 3.9%.
Table 4⇓ shows the ORs with 95% confidence intervals for potential predictive factors of the occurrence of 2-day MACE. Results from the single-factor analyses showed that only dissection/abrupt closure (P<0.001) and country (P=0.005) were significant predictors of the occurrence of 2-day MACE. The presence of the most severe kind of lesion, type C, was borderline with respect to its significance as a predictor (P=0.06). As shown in Table 4⇓, contrast medium, sex, age, clinical condition of the patient, and stenting condition were not significant predictors for the occurrence of 2-day MACE. The multifactorial analysis gave similar results.
Similarly, in the per-protocol population, no statistically significant differences between the 2 groups were observed (P=0.27) with regard to rehospitalization due to MACE during the period between the 2-day and 1-month follow-up. The number of patients hospitalized for each event in each group (iodixanol/ioxaglate) are as follows: cardiac death (2/0), stroke (1/0), Q-wave MI (2/0), NQWMI (1/1), CABG (2/1), and Re-PTCA (6/7). In total, 23 patients (1.6%) were rehospitalized during this time period, 2.0% in the iodixanol group and 1.3% in the ioxaglate group. These events were not validated by the end point classification committee. Combining MACE noted at 2-day follow-up and those occurring between the 2-day to 1-month follow-ups showed no statistically significant difference between the iodixanol and ioxaglate groups (6.3% versus 5.2%; P=0.42).
The percentage of patients reporting ≥1 adverse event during the 2-day follow-up, including MACE, was similar in the 2 groups (27.5% versus 27.6%; Table 5⇓). Chest pain was the most common event reported; this was followed by hypotension, hematoma at the puncture site, nausea or vomiting, and bradycardia. As seen in Table 5⇓, patients randomized to the ioxaglate group had an increased risk (P=0.002) of experiencing adverse events that were classified by the investigators as having a certain relationship to the contrast medium. These were mainly nausea, vomiting, and rash. When pooling contrast media–related adverse events and events having an uncertain relationship to contrast media, the overall adverse drug reaction frequencies were 9.9% in the iodixanol group and 11.9% in the ioxaglate group (P=0.23). Among the most common adverse drug reactions with an uncertain relationship to contrast medium were hypotension, bradycardia, chest pain, and hematoma.
Patients randomized to receive the ionic contrast medium ioxaglate had an increased risk of developing hypersensitivity reactions (P=0.007) (Table 5⇑). A total of 23 patients had such reactions, 5 (0.7%) in the iodixanol group and 18 (2.5%) in the ioxaglate group.
Although nonionic contrast media are better tolerated than ionic contrast media, their potential merits and disadvantages in patients undergoing interventional procedures have been a controversial subject.1 2 3 4 5 6 7 Studies in vitro have demonstrated differences in vascular biocompatibility between the iso-osmolar nonionic contrast medium iodixanol and the ionic contrast medium ioxaglate.9 10 11 12 13 14 15 16 17 18 Although both contrast media display anticoagulant and antiplatelet properties in this setting, such effects are more marked with ioxaglate.19 20 However, iodixanol is the most biocompatible, particularly with regard to cultured endothelium21 and vascular smooth muscle.22 Data on collagen- or tissue factor–induced arterial thrombus formation in native blood and blood including clinically relevant doses of aspirin and heparin were recently published.23 24 They demonstrate no significant effects of iodixanol, ioxaglate, or the nonionic contrast medium iohexol on thrombin-driven (tissue factor) or platelet-driven (collagen) thrombus formation. Previously published clinical studies on intravascular thrombotic events and outcomes were limited, and the patient populations studied were relatively small at the time the trials were conducted. A recent exception is the study reported by Schräder et al.25 Therefore, it was necessary to perform a large-scale, blinded, well-controlled study, with objective “hard” end points.
The major finding of this multicenter study was that no difference was found in MACE between the nonionic iso-osmolar and the ionic low-osmolar contrast media after both 2 days and 1 month. These results agree with the recently reported single-center trial performed by Schräder et al.25 In this randomized study comparing the nonionic iomeprol (Iomeron, Bracco-Byk-Gulden) to the ionic ioxaglate in a large series of 2000 patients, no differences were found with regard to angiographic or clinical end points. Fewer than 20% of their patients had unstable angina, and glycoprotein IIb/IIIa receptor inhibitors were not used. However, the present study’s results differ from those reported in a single-center study by Grines et al.4 They compared ioxaglate and the low-osmolar, nonionic contrast medium iohexol (Omnipaque, Nycomed Imaging AS) in 211 patients with acute MI and unstable angina undergoing PTCA. Their conclusion was that the use of the ionic contrast medium ioxaglate reduced the risk of ischemic complications, both acutely and at 1-month follow-up. Results from recently published studies on the effects of iohexol, iodixanol, and ioxaglate on vessel-wall biocompatibility and on arterial thrombus formation23 24 do not support the findings of Grines et al.4 However, it is difficult to directly compare the results from the present study with those from Grines et al,4 because many essential differences exist between the studies. We think that the main reasons for discrepancies in results are that the present study included a much larger patient population (it was a multicenter study, with ≈7 times as many patients), patients with acute MI were excluded, adjuvant anticoagulation was strictly controlled, and intracoronary stenting was done according to current practice. Finally, the nonionic contrast medium was different in the 2 studies; the present study used the new iso-osmolar iodixanol.
With regard to predictive factors for the occurrence of MACE, only the procedure-related factors dissection and abrupt closure were significant, in addition to country (P=0.45 for contrast medium). Regarding the procedural aspects, we noted no difference between the 2 groups in thrombus, abrupt closure, or dissection. However, the identification of thrombus is based on angiographic assessment, and it is well known that angiography has a poor sensitivity for the detection of thrombus.
This study reflects the present situation in Western Europe regarding the implantation of intracoronary stents. On average, 60% of the patients were stented, but the country-to-country variation was large (from 36.0% to 76.9%). This variation is due to international differences in reimbursement systems rather than to angiographic indications.
Our results confirm that hypersensitivity and adverse drug reactions related to contrast medium injection are less frequent after the administration of nonionic contrast medium26 27 28 29 30 31 (Table 5⇑). These reactions were independent of the volume of contrast media injected, which was similar in both groups (Table 2⇑).
Schräder et al25 concluded that nonionic contrast medium minimized the risk of allergic reactions in patients undergoing intervention, and Gertz et al31 recommended that nonionic contrast medium should be used in patients undergoing cardiac angiography to minimize allergic reactions, especially in patients with known allergies. This recommendation is now also valid for patients undergoing PTCA.
The implantation of intracoronary stents is increasing, both in Europe and worldwide. During the planning phase of the study, we anticipated an average stenting rate of ≈40% of patients. The results showed an ≈20% higher rate of stenting, and this fact might have contributed to the lower MACE rates reported than was expected. The study was planned with a 6% anticipated MACE rate. However, the lower rate of 4.3% seen at the completion of the study did not affect the study’s ability to detect a difference between the contrast media groups, because the observed contrast group OR of 1.2 is well within the confidence limits specified in the study protocol.
Only the 2-day MACE were evaluated by the independent end point classification committee, and this should be considered when looking at the overall results, which combine the MACE reported after 2 days with those reported after 1 month. In total, 1541 patients were enrolled, but 1411 were included in the per-protocol population. However, the fact that the intention-to-treat population results were in accordance with the per-protocol population results regarding MACE indicates that the exclusion of these patients had no influence on the results.
In total, 32 centers participated in the trial, and the individual centers enrolled 2 to 112 patients. The disparity between the number of enrolled patients could have influenced the final results, but due to stratification by center, the comparison between the 2 contrast medium groups was unbiased.
The use of glycoprotein IIb/IIIa receptor inhibitors was not the current practice at the enrolling centers, and it was not allowed as a preparation to intervention. Only 23 patients (1.6%) received this type of medication during or after PTCA.
No statistically significant differences were found between the iso-osmolar, nonionic contrast medium iodixanol and the low-osmolar, ionic contrast medium ioxaglate on the occurrence of MACE, either at the 2-day follow-up or during the 1-month follow-up period. In addition, the rate of abrupt closure was similar in both groups.
Thus, we can conclude that in patients, including those with both stable and unstable angina, receiving iodixanol, no increase in major cardiac complications occurs when compared with those patients given ioxaglate. Furthermore, hypersensitivity and adverse drug reactions classified as certainly related to contrast medium occurred significantly less frequently in patients receiving iodixanol.
France: M. Bertrand, Hôpital Cardiologique, Lille; X. Faverau, Center Médico-Chirurgical, Le Chesnay; P. Aubry, Hôpital Bichat, Paris; J.L. Guermonprez, Hôpital Broussais, Paris; M.-C. Morice, Hôpital Privé de Massy, Massy; P. Besse, Hôpital Haut Leveque, Pessac; G. Grollier, Center Hôspitalier Universitaire de Caen, Caen; P.-D. Crochet, Hôpital Laënnec, St. Herblain; A. Cribier, Hôpital Charles Nicolle, Rouen; P. Dupouy, Hôpital Henri Mondor, Créteil; J. Boschat, Hôpital de la Cavale Blanche, Brest; K. Khalifé, Hôpital Notre Dame de Bon Secours, Metz.
The Netherlands: H. Suryapranata, Ziekenhuis De Weezenlanden, Zwolle; J.J.R.M. Bonnier, Catharina Ziekenhuis, Eindhoven.
Belgium: G. Heyndrickx, Onze Lieve Vrouw Ziekenhuis, Aalst; C. Hanet, Cliniques Universitaires St Luc, Bruxelles; F. Van den Branden, Middelheim Ziekenhuis, Antwerpen; M. Vrolix, St. Jan Ziekenhuis, Genk; J.H. Piessens, Universitair Ziekenhuis Gasthuisberg, Leuven; Y. Taeymans, Universitair Ziekenhuis, Gent; V. Legrand, Center Hospitalier Universitaire, Liège; J. Dekeyser, Hôpital Civil, Jumet; P. Lafontaine, Clinique St Jean, Bruxelles.
Spain: C. Macaya, Hospital Universitario San Carlos, Madrid; E. Garcμa-Fernández, Hospital General Gregorio Marañón, Madrid; E. Esplugas and J. Mauri, Hospital de Bellvitge Princeps D’Espanya, Barcelona; A. Betriu and A. Serra, Hospital Clinic I Provincial, Barcelona; A. Iñiguez, Clínica Nuestra Señora de la Concepción, Madrid; C. Morís, Hospital General de Asturias, Oviedo; T. Colman, Hospital Marqués de Valdecilla, Santander; M. Gómez-Recio, Hospital de la Princesa, Madrid.
Sweden: L. Grip, Sahlgrenska Sjukhuset, Göteborg.
We would like to thank all the investigators for their efforts during their participation in this multicenter trial. We also wish to thank Mr Trond Haider for the statistical analysis.
- Received March 17, 1999.
- Revision received July 7, 1999.
- Accepted July 28, 1999.
- Copyright © 2000 by American Heart Association
Grines CL, Schreiber TL, Savas V, Jones DE, Zidar FJ, Gangadharan V, Brodsky M, Levin R, Safian R, Puchrowicz-Ochocki S, Castellani MD, O’Neill WW. A randomized trial of low osmolar ionic versus nonionic contrast media in patients with myocardial infarction or unstable angina undergoing percutaneous transluminal coronary angioplasty. J Am Coll Cardiol.. 1996;27:1381–1386.
Gasperetti CM, Feldman MD, Burwell LR, Angello DA, Haugh KH, Owen RM, Powers ER. Influence of contrast media on thrombus formation during coronary angioplasty. J Am Coll Cardiol.. 1991;18:443–450.
Chronos NA, Goodall AH, Wilson DJ, Sigwart U, Buller NP. Profound platelet degranulation is an important side effect of some types of contrast media used in interventional cardiology. Circulation.. 1993;88:2035–2044.
Eloy R, Corot C, Belleville J. Contrast media for angiography: physicochemical properties, pharmacokinetics and biocompatibility. Clin Mater. 1991;7:89–197.
Hathcock J, Sakariassen K, Turitto V. Tissue factor activity in rat vascular smooth muscle cells exposed to ionic and non-ionic contrast media. Thromb Haemost.. 1997;78:828.
Sakariassen KS, Buchmann M, Hamers MJ, Stormorken H. Iohexol, platelet activation, and thrombosis, II: iohexol-induced platelet secretion does not affect collagen-induced or tissue-factor-induced thrombus formation in blood that is anticoagulated with heparin and aspirin. Acta Radiol.. 1998;39:355–361.
Schräder R, Esch I, Ensslen R, Fach WA, Merle H, Scherer D, Sievert H, Spies HF, Zeplin HE. A randomized trial comparing the impact of a nonionic (iomeprol) versus an ionic (ioxaglate) low osmolar contrast medium on abrupt vessel closure and ischemic complications after coronary angioplasty. J Am Coll Cardiol.. 1999;33:395–402.
Gertz EW, Wisneski JA, Miller R, Knudtson M, Robb J, Dragatakis L, Browne KF Jr, Vetrovec G, Smith SC Jr. Adverse reactions of low osmolality contrast media during cardiac angiography: a prospective randomized multicenter study. J Am Coll Cardiol.. 1992;19:899–906.