ABC | Volume 114, Nº2, February 2020

Original Article Vieira et al. Evaluation myocardial ischemia with iFR Arq Bras Cardiol. 2020; 114(2):256-264 However, whether these results from randomized studies, suggesting that iFR can be used as surrogate for FFR in percutaneous interventions in CAD, can be transposed to clinical practice is still uncertain. Besides, factors like the costs of equipment, inadequate reimbursement, the interventional cardiologist preference, signs and symptoms reported by patients, and the costs and risks associated with adenosine treatment may limit the use of both methods. The use of iFR in a routine manner in patients with multi‑vessel diseases and in non-culprit lesions in acute myocardial infarction (AMI) patients still need to be investigated. 13 The present study aimed to evaluate the additional contribution of iFR to the therapeutic decision-making. The iFR was used in coronary disease patients in which the correlation between obstructive atherosclerotic disease and myocardial ischemia had not been clearly established by other conventional diagnostic methods. Methods The study was approved by the ethics committee of Marcilio Dias Naval Hospital (approval number CAAE: 58741716.0.000.5256). We studied patients older than 18 years, of both sexes, consecutively referred for percutaneous treatment betweenMay 2014 and March 2018. All patients were referred for invasive investigation of myocardial ischemia and decision-making process by the Heart Team, composed by interventional cardiologists, clinical cardiologists and cardiovascular surgeons. All patients with moderate (41-70% stenosis) or severe (71%-90%) stenosis according to coronary angiography were included. In all these patients there were doubts about the degree of obstruction, determined by coronary angiography, and its correlation with the presence of ischemia determined by non-invasive methods including ergometric test, myocardial scintigraphy and stress echocardiography. The study population was composed of a wide variety of patients – patients with suspected or confirmed diagnosis of stable CAD but inconclusive diagnosis of myocardial ischemia using non-invasive methods; non-ST-elevation myocardial infarction patients in which the culprit artery had been treated, and invasive functional analysis had been performed in another coronary vessel with moderate-to-severe lesion by angiography; ST-elevation myocardial infarction patients in which invasive functional analysis of moderate-to-severe non-culprit lesion had been performed at least 5 days after the acute event. The iFR was performed using the Volcano S5 Imaging System (San Diego, California, USA). The 0,014” Primewire Prestige® Pressure Guide Wire was used in 2014, and the 0,014” Verrata Pressure GuideWire, substitute for the previous version, used in 2015. A guiding catheter was used to advance the guide wire through the lesion. 14-16 All procedures were performed according to good practice guidelines for iFR measurements, as follows – the 0.014’’ guidewire was stabilized before handling by infusion of 0.9% saline until completion of the circuitry where the catheter was packed, and connection of the catheter to the console; during this process, the device was kept in stable position until it was recognized by the console software. After the guide wire was introduced into the catheter, it was externalized through the proximal coronary segment, and the guide pressure equalized using a transducer. The transducer guide was then positioned about 3 cm below the lesion. 15 Also, guide pressure equalization was confirmed at the end of each measure to ensure its stability. 16 To confirm the stability of the results, three consecutive measures were performed for each lesion; in case of diverging values, the lowest value was considered for analysis. Intracoronary nitroglycerin (200 μg, bolus) was administered before the measures were performed. 16 The iFR was considered positive for myocardial ischemia 0.89 or less. 12 Statistical analysis Categorical variables were described as numbers and percentages. Age (continuous variable) was described as mean and standard deviation, and as minimum, median and maximum values. Normality of the variable age was confirmed by the Shapiro-Wilk test (p = 0.3663). Distribution of the variable iFR was not tested for normality, and described as median and interquartile range. A logistic regression was initially performed using the elastic net, 17 which is a variable selectionmethod that identifies strongly correlated predictors. This method is particularly useful when the number of predictors (P) is much bigger than the number of observations (n). In this model, the requirement of a stent was the outcome variable, and the independent variables were age, sex, comorbidities (such as systemic arterial hypertension, diabetes mellitus, dyslipidemias, smoking, family history, obesity and previous AMI). Two logistic regression models were built using the variables selected by the elastic net. In addition, we used a non-parametric classification tree, 18 which is useful to detect possible interactions between predictors and provide easily interpreted visual information. The end nodes show the bar graph for the variable ‘stenting’. Additionally, the ROC curve was used to evaluate sensitivity and 1-speciticity of the iFR cut-off, established by the classification tree. Box plots 19 were constructed to depict the distribution of the iFR values for moderate and severe stenoses, considering the use of stents. Statistical calculations were performed using the R package. 20 The partykit package of the R software was used for construction of the classification tree. 21,22 A p-value < 0.05 was considered statistically significant. Results Characteristics of the patients The iFR was used for assessment of 96 stenotic lesions of 52 patients, with a mean of 1.85 lesions/patient. Median iFR was 0.93 (0.855–0.97); 56 of them were classified as moderate stenosis (58.3%) and 40 of them as severe (41.7%) stenosis. Figure 1 shows the study flowchart. Thirty percent of the lesions were treated with stent placement, and in 6.2% of them, despite the presence of ischemia confirmed by functional analysis, the first therapeutic choice was other than stent placement – revascularization surgery due to the 257