2.1 Patient
This study was a single-center retrospective observational study. This study retrospectively analyzed patients who underwent CABG at our hospital from May 2018 to June 2019. 81 patients (77.9%) were male, with a mean age of (61.18±8.86) years. All patients with coronary artery disease were free of comorbid heart valve disease or other organic heart disease. All patients were bypassed using the left internal mammary artery with or without saphenous vein bypass. Inclusion criteria: (1) preoperative transthoracic bilateral internal mammary artery ultrasonography showed no significant stenosis or occlusion; (2) non-emergency surgery. Exclusion criteria: LIMA was extensively diseased, had obvious sign of hematoma, or was damaged in any way that could potentially adversely affect flow. We finally included 104 patients. The primary endpoint event in this study was cardiac death within 18 months after surgery.
2.2 The ultrasound of IMA
Color Doppler ultrasound diagnostic instrument (Philips CX50,Philips Medical Devices Group, Netherlands) equipped with a linear array probe at 7-10 MHz was used. All patients perfected bilateral internal mammary artery ultrasonography before the procedure to clarify the presence of stenosis or plaque formation in the vessel (discarded for those with stenosis or plaque). The diameter and flow velocity parameters of the left internal mammary artery were also obtained (Figure 1A).
The patient was asked to lie in a supine position with the shoulders elevated and the head slightly tilted back so that the left side of the neck and shoulder could be fully exposed. The probe was placed in the left supraclavicular fossa to make a transverse cut and probe the long axis of the subclavian artery, then the probe was rotated by 90% and slid inward and outward to reveal the beginning of the internal mammary artery on the opposite side of the inferior wall of the subclavian artery, that is, the beginning of the vertebral artery. The diameter of the left internal mammary artery was measured at the 3rd intercostal space cross-section of the chest wall. Flow velocity was measured in a long-axis view of the internal mammary artery in the third intercostal segment (Figure 1A). All examinations were performed by the same senior ultrasonographer.
2.3 Echocardiography
Color Doppler ultrasound diagnostic instrument (Philips CX50,Philips Medical Devices Group, Netherlands) equipped with a cardiac probe at 1-5 MHz was used. We measured overall left ventricular systolic function using the biplane Simpson's method according to the American Society of Echocardiography (ASE) guidelines [6]. The left ventricular systolic function was divided into four grades, male: 52% ~ 72% (normal range), 41% ~ 51% (mild abnormality); 30% ~ 40% (moderate abnormality); <30% (severe abnormality); female: 54% ~ 74% (normal range), 41% ~ 53% (mild abnormality), 30% ~ 40% (moderate abnormality), <30% (severe abnormality).
2.4 TTFM
Intraoperative blood flow and pulsatility index in the LIMA bridge vessels were measured using TTFM (Medistim VeriQ, Oslo Norway) [7] (Figure 1B). TTFM parameters: (1) mean flow (Q, mean graft flow, MGF), which is the mean blood flow in the bypass graft vessel; (2) PI value, which is the ratio of the difference between the maximum and minimum blood flow in the graft vessel to the mean flow [PI = (Qmax - Qmin)/ Qm].
After all bridge vessels are anastomosed, they are neutralized with fisetin. We wait for the circulation to stabilize. We then select the appropriate size ultrasound probe (2 or 3) according to the bridge vessel diameter and place the bridge vessel into the probe near the anastomosis for direct measurement. For unsatisfactory measurements, the surrounding connective tissue can be removed, skeletonized, and measured again. If 118 necessary, the anastomosis and the graft vessel need to be repeatedly examined or even reanastomosed and measured again. Criteria for determining whether the blood flow in the bridge vessel is satisfactory: (1) satisfactory coupling (>50% or more) when measuring the bridge vessel in the internal mammary artery; (2) TTFM shows a stable and reproducible flow waveform pattern; (3) the average flow red line is stable at the plateau period after recording the above data.
2.5 CABG
Patients were extubated and CABG was performed under general anesthesia. All patients underwent CABG by median thoracotomy. We used low-frequency electroknife to free the internal mammary arteries bilaterally, and after systemic heparinization, we dissected the distal internal mammary arteries and protected them with poppyine wet gauze. The corresponding target vessel anastomosis was then completed.
2.6 Statistical analysis
Data are expressed as either the mean ± SD, median and interquartile range (25th and 75th percentiles), or frequency (%). Correlation was measured used Spearman correlation analysis. We performed a Cox proportional-hazards regression modelling to study independent predictors of death. Receiver operating characteristic curve analysis was used to determine the optimal cutoff points for LIMAV, MGF and LVEF to predict death. We use Comparison of ROC curves to compare the values of MGF, MGF+LIMAV, and MGF+LIMAV+LVEF in prediction.. A P value of <0.05 was considered significant. All calculations were processed using the SPSS software package and Medcal software package for Macintosh (SPSS, Chicago, IL, USA).