This work is an observational and prospective study that evaluates the development of CIN in 2 cohorts of consecutive patients with ACS (unstable angina or acute myocardial infarction [AMI])4 referred for invasive coronary procedures in 2 institutions: the University Hospital of Nuestra Señora de Candelaria (Santa Cruz de Tenerife, Spain) and the University Hospital of the Canary Islands (San Cristóbal de La Laguna, Tenerife, Spain). The type of angiography was performed according to the operator's criteria. Patients included were those with consecutive ACS with indication of invasive coronary angiography who were not on renal replacement therapy and who did not present an AMI with ST segment elevation of less than 12h of evolution since the performance of an emergent coronary angiography does not allow a procedure regulated on multiple occasions.

The Ethical and Clinical Research Committee of the University Hospital Nuestra Señora de Candelaria and the University Hospital of the Canary Islands approved the protocol of this study, since it complies with current ethical and legal regulations.

• -

  Indication of invasive coronary angiography by ACS with or without percutaneous coronary intervention.

• -

  Informed consent.

• -

  Patients <18 years.

• -

  Previous renal replacement therapy.

• -

  Women with possibilities of being pregnant.

• -

  Allergy to iodinated contrast previously known, that cannot receive premedication.

• -

  Exposure to iodinated contrast in the previous 10 days.

• -

  Previous myocardial revascularization surgery.

• -

  AMI with ST-segment elevation of <12h of evolution.

• -

  Cardiogenic shock.

• -

  Inability to understand the nature of the study or medical or social disability that may interfere with the collection of data or appropriate follow up.

• -

  Inclusion in other clinical trials or registries.

Patients received treatment for ACS according to standard clinical practice. If they had a baseline creatinine ≥1.59mg/dL, patients received normal saline (0.9% sodium chloride) at a rate of 1mL/kg/h during the 12h before and 24h after the procedure, unless there were contraindications.20

The CCA was performed according to the following recommendations: a minimum of 6 projections for the left coronary artery and a minimum of 3 projections for the right coronary artery. However, the final number of projections for a correct assessment of the coronary arteries or planning of percutaneous coronary intervention was left to the discretion of the operator. The volume of contrast used was from 8 to 4mL/s for the left coronary artery and from 5 to 3mL/s for the right coronary artery. The RCA was performed according to recommendations. In order to obtain optimal quality images by means of RCA, it was decided to perform a 5.8s rotation in the left coronary artery and the volume injected was 14–2.5mL/s; for the right coronary artery, it was performed a 4s rotation and an injection of 10–2mL/s.19

The procedure of percutaneous coronary intervention ad hoc was left to the discretion of the medical team and it was performed according to the usual clinical practice. The procedure of ventriculography was also done at the discretion of the operator. If it was considered adequate, it was injection of 45 to 15mL/s or from 36 to 12mL/s depending on the degree of renal dysfunction. A third-generation contrast agent was used: iomeron 350 (Iomeprol, Bracco Corporate, Milan, Italy) and, to standardize the administration of contrast, it was used the ACIST CVi® robotic contrast injector (ACIST Medical Systems, Eden Prairie, MN, US). The choice of coronary catheters was left to the discretion of the operator. The acquisition of the images was done according to the recommended practice for the RCA.19

Baseline samples were obtained to measure baseline serum creatinine levels (with the Jaffé method) and to calculate eGFR using the CK-EPI formula. Samples were also obtained after 48–72h of the procedure to evaluate changes in renal function.29,30 In the case of patients who were discharged before 48h after the procedure, the patient came to the outpatient clinic for blood sampling in the period of time specified in the study. Likewise, we collected data related to the procedure, clinical variables, angiographic (location and number of coronary vessels with stenosis >50%), radiation, analytical data and also the length of hospital stay (days).

Patients were visited at day 30 to evaluate clinical variables. The information was collected prospectively and entered into a specifically designed computerized database.

Sample size was calculated using GRANMO software (Institut Municipal d’Investigació Mèdica, Barcelona, Spain). The sample size was calculated to demonstrate a reduction from 25% in the cohort of patients analyzed with CCA to at least 10% in the cohort of patients analyzed with RCA (relative risk ACR/ACC≤0.4), with a ratio of patients between the groups of 1:1 and a proportion of losses in the follow-up of 3%. Using a χ2 test for 2×2 tables, with an alpha risk of 0.05 and a beta risk of less than 0.2 (statistical power of 80%) in a bilateral contrast, the calculated sample size was 116 subjects per arm to detect statistical difference between 2 proportions (232 patients in total).31

Statistical analysis was performed with SPSS Statistics 24.0 software (SPSS Inc., Chicago, IL, United States). Two tails p<0.05 were considered statistically significant. The normality of the continuous variables was explored by the Kolmogorov–Smirnov test. The categorical variables were expressed as percentage. The categorical variables were compared using the Fisher exact test or the chi-square test, as appropriate. Variables with values with a normal distribution were expressed as mean (standard deviation), and were compared using t tests for 2 samples; the continuous variables without a normal distribution are expressed as median (interquartile range) and were analyzed using the Wilcoxon nonparametric test (rank-sum).

Binary logistic regression models were done to establish the independent predictors of CIN. A univariate exploratory analysis was performed, introducing in the model the covariates that had a p<0.10. Due to the small number of events in the primary end point (CIN), only 2 variables of important clinical relevance were added to the model: the presence of previous renal dysfunction (eGFR<60mL/min/1.73m2) and the execution of percutaneous coronary interventionism.2,3 The final model included the following variables: RCA, diabetes mellitus, contrast volume ≥300mL, dose×area product ≥50Gycm2, percutaneous coronary intervention and eGFR <60mL/min/1.73m2. The results are reported as odds ratio (OR) with a 95% confidence interval (95% CI).

Finally, to evaluate the association between the RCA (exposure variable) and the CIN (primary endpoint variable), a Poisson regression was performed introducing the independent predictor variables of CIN in the model. Results were reported as a relative risk (RR) adjusted with a CI of 95%.

The baseline characteristics of both groups are presented in Table 1. Patients from the RCA group were older (64.0±11.8 vs. 59.7±12.1 years, p<0.001) with a higher percentage of women (44.8% vs. 17.6%, p<0.001) and higher percentage of hypertension (71.6% vs. 58.0%, p=0.03). Likewise, the RCA group had a lower proportion of smokers (p=0.014) and less percent of AMI as the clinical presentation (63.8 vs. 76.5%, p=0.034).

Table 2 shows the medications received. There were no significant differences in medical therapies received with a demonstrated benefit in ACS.

The data comparing the angiographic and radiation parameters between both groups are presented in Table 3.

Patients undergoing RCA had a less extent of coronary disease (p=0.012) and the percent of patients receiving coronary intervention was lower than patients treated with CCA (p<0.001).

Patients treated with RCA received less amount of contrast during the procedures than the CCA patients (113±92 vs. 169±103mL, p<0.001). The amount of contrast was less in RCA than CCA during diagnostic procedure (52±24 vs. 85±56mL, p<0.001) and during the combined diagnostic and therapeutic procedures (174±64 vs. 205±98mL, p=0.049). Likewise, there was also a lower number of acquisitions in the RCA than CCA group (11±9 vs. 17±9, p<0.001).

Of note, there was a tendency to shorter time of escopia with RCA than CCA (8.2±7.3 vs. 10.2±8.9min; p=0.057) and also a shorter duration of the procedure (52±26 vs. 60±23min; p=0.054).

Table 4 shows the analytical variables data of and results of CIN (primary assessment variable), also shown is the data on CIN using AKI-IC criteria, and eGFR criteria.

Baseline eGFR was lower in RCA than CCA group (75.58±24.87 vs. 86.14±27.32mL/min/1.73m2; p=0.001). After the procedure, the eGFR was still lower in RCA than CCA group (70.24±23.62 vs. 79.44±25.23mL/min/1.73m2; p=0.003).

Regarding CIN, the primary end point, it was found that in the RCA cohort CIN was less frequent than CCA (4.3% vs. 22.7%, p<0.001). Likewise, it was also observed that patient from the RCA group had a lower percentage of CIN according to either AKI-IC criteria (2.6% vs. 12.6%, p=0.006) or eGFR criteria (3.4% vs. 16.8%, p=0.004).

Results from logistic regression analysis to detect independent predictors of CIN (increase in serum creatinine ≥0.5mg/dL or ≥25%), ACR shows to be a protective factor (OR: 0.154; 95% CI: 0.053–0.445, p=0.001). Likewise, both diabetes mellitus (OR: 3.378, 95% CI: 1.395–8.197, p=0.007) and contrast volume administrated ≥300mL (OR: 4.566, 95% CI: 1.346–15.385; p=0.015) are risk factors for CIN (see Table 5).

Finally, the Poisson regression analysis reveals that RCA is associated with lower frequency of CIN defined as creatinine increase ≥0.5mg/dL or ≥25% (adjusted RR: 0.868, 95% CI: 0.794–0.949; p=0.002).

Clinical variables were not different in the two groups. Table 6 shows the data on clinical assessment variables.

First, this study is an observational and prospective analysis (cohort study) with the biases inherent to this type of study. However, it is the first study to evaluate the impact of RCA on the development of the CIN. Second, the absence of randomization led to a non-homogeneous distribution of baseline variables (age, sex, eGFR, etc.) and therapies (percentage of implanted stents, etc.) between the two study groups; this could have influenced the incidence of the primary assessment variable (CIN). However, after adjusting for confounding factors, the results are favorable to the RCA. Third, our data refer to the population of the Canary Islands and, therefore, cannot be fully extrapolated to other geographical areas. Fourth, the procedure protocol of the CCA may not be identical in the different units of interventional cardiology. However, we consider that the CCA protocol may be considered conservative in terms of contrast administration, limitations in the performance of ventriculographies and the initial number of angiographic projections for each coronary artery and, for this reason, the final volume of contrast administered in the CCA group should not be overestimated. Fifth, the clinical follow-up of the patients was limited in time (one month), which may be related to the absence of differences in the clinical assessment variables between the study groups.