Introduction

Coronary artery bypass grafting (CABG) surgery is one of the most commonly performed cardiac procedures in adults all over the world. Arterial grafts were preferred to venous grafts because their long term patency rates are higher and early graft dysfunction is rare [₁,₂,₃,₄,₅,₆,₇]. Among the arterial grafts, left internal mammarian artery (LIMA) is the most commonly preferred arterial graft; because of its convenience to prepare, rareness of complications and excellent long term patency rates [₂,₃,₄,₅,₆]. Because of the widespread atherosclerosis, stenosis in peripheral arteries is frequently observed in coronary artery disease patients, collateral circulation between LIMA and left inferior epigastric artery (LIEA) gains importance.

If IMA is used as a graft in patients with chronic ischemia, severe ischemic findings may occur at early postoperative period [₈,₉,₁₀,₁₁,₁₂].

Although many studies exist in the literature evaluating the flow patterns of LIMA grafts using color Doppler ultrasonography (US) in pre-, peri- and postoperative periods [₃,₄,₅,₁₃,₁₄,₁₅,₁₆,₁₇,₁₈,₁₉,₂₀,₂₁,₂₂,₂₃,₂₄,₂₅], the changes in the arteries other than IMA (vertebral, subclavian, inferior epigastric and brachial) were not investigated. Our purpose in this study is to evaluate the early and late postoperative changes in flow patterns of LIMA and arteries in close relationship with LIMA by color Doppler US.

Material And Methods

The study group consisted of 48 consecutive patients which had coronary artery bypass surgery and pediculated LIMA anastomosis to LAD at Trakya University Department of Cardiovascular Surgery. The patients that had valvular surgery in addition to CABG. Patients with postoperative neurologic complications, patients in which LIMA and right internal mammarian artery (RIMA) was used or LIMA used as a free graft, and patients that had preoperative left subclavian artery (LSA) stenosis and emergency operation performed before color Doppler US evaluation were not included in the study (n=16). The study was performed on the remaining 32 patients. LIMA was dissected by the same surgeon (TE) using electrocauter and was anastomosed by using 7/0 polypropylen suture with the continiuous technique by the same surgeon (ED). In all LAD anastomosis patients, the stenosis was after the first septal branch region with a degree of 70% or more. Color Doppler US, imaging of the LIMA, RIMA, LSA, left vertebral artery (LVA), left brachial artery (LBA) and LIEA was performed on all patients of the Trakya University Department of Radiology department. Color doppler US was performed to all patients by the same radiologist (NT). The protocol was approved by the Ethical committee of the Trakya University Research Hospital.

This doppler based imaging method was performed through the second intercostal space in supine position for IMA. A computerized color doppler ultrasound scanner (Sonoline Ellegra Advanced, Siemens, Germany) equipped with a 7.5-MHz transducer was used for all studies. The doppler probe was placed directly on the patient's skin after the application of a commercial ultrasonic gel and patient positioned to maintain an angle as close as possible to 60 degrees to the axis of blood flow.

To find the origin of the LIEA, gray-scale imaging was used to locate the parent vessel; the external iliac artery. The LIEA joins the external iliac artery about 1 to 2 cm above the inguinal ligament at an angle of approximately 90 degrees. Because the LIEA has lower flow velocity than the external iliac artery, its location was more easily identified by decreasing the velocity scale. The ultrasound transducer was manipulated to capture a longitudinal or cross- sectional image of the LIEA in its greatest diameter. The image was frozen during systole and the proximal internal diameter of the artery was measured in millimeters.

We measured diameters from LIMA, RIMA, LSA, LVA, LBA, LIEA ,RIMA and blood flow velocity (cm/s) at peak systolic flow velocity (Vmax), end-diastolic flow velocity (Vmin), resistive index (RI), pulsatility index (PI) obtained from each patient. These studies were done preoperative, early postoperative (7-10 days) and late postoperative periods (3 months).

Statistical Analysis

Results

The study group comprised of 32 patients (29 male and 3 female of ages between 38-68 yrs; mean: 52 yrs; SD; 7.8). The mean number of grafts used was 2.5±0.9, body surface area was 1.8±0.2 m2, cardiopulmonary bypass time (CPBT) was 92.1±29.9 minutes, cross clamp time (CCT) was 52.2±19.9 minutes. No significant difference was detected in ankle-brachial index (ABI) values from both extremities at preoperative and late postoperative periods (For both extremities, preoperative 1.2± 0.2 and postoperative 1.2± 0.2). 6 patients had diabetes mellitus and 26 patients had a history of smoking. No significant difference was detected in the diameter of LIMA between β-blocker using (metoprolol 2x50 mg/day) 23 patients and the remaining 9 patients in late postoperative controls (p=0.644).

It was shown that; while Vmax, PI, RI and diameter of LIMA decreased in postoperative period, Vmin increased gradually (Table 1 ) (Fig 1).

Figure 1

Table 1: Color Doppler US examination results of LIMA

LIMA= left internal mammary artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.* 2-Related Samples test

Figure 2

Figure 1: The changes in LIMA flow pattern (1A: preoperative, 1B: postoperative 3rd month).

Figure 3

While Vmax, PI, Vmin and diameter of LVA decreased significantly in early and late postoperative period compared to preoperative levels, RI increased (Table 2 ) (Fig 2).

Figure 4

Table 2: Color Doppler US examination results of LVA

LVA = left vertebral artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.
*2-Related Samples test

Figure 5

Figure 2: The changes in LVA flow pattern (2A: preoperative, 2B: postoperative 3rd month).

While Vmax and Vmin values of LSA decreased significantly compared to preoperative levels, PI, RI and diameter did not differ significantly (Table 3 ).

Figure 6

Table 3: Color Doppler US examination results of LSA

LSA= Left subclavian artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.* 2-Related Samples test

While Vmax, PI and Vmin values of LIEA decreased significantly, RI and diameter did not change significantly (Table 4 ) (Fig 3).

Figure 7

Table 4: Color Doppler US examination results of LIEA

LIEA= left inferior epigastric artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.*2-Related Samples test

Figure 8

Figure 3: The changes in LIEA flow patterns (3A; preoperative, 3B; postoperative 3rd month)

In evaluation of RIMA; it did not show statistically significant difference in parameters other than reduction in PI and RI values (Table 5).

Figure 9

Table 5: Color Doppler US examination results of RIMA

RIMA= right internal mammarian artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.
* 2-Related Samples test

LBA did not show statistically significant difference in diameter and flow pattern compared to preoperative evaluation (Table 6).

Figure 10

Table 6: Color Doppler US examination results of LBA

LBA= left brachial artery, Vmax = peak systolic flow velocity, Vmin = end-diastolic flow velocity, PI = pulsatility index, RI = resistive index.* 2-Related Samples test

Discussion

Although color Doppler US is a non-invasive, reproducible method and it maintains quicker evaluation of IMA compared to angiographic investigations, angiography is a valuable diagnostic method in showing the localization and degree of stenosis on graft. Change in anatomic localization of LIMA graft in postoperative period results in difficulties of evaluation of flow pattern and diameter by color Doppler US. Nasu et al. [₁₃] stated that there is a difference in flow and diameter between proximal, middle and distal portions of LIMA, so that measurements from different areas may end-up with misleading results. Thus, in our study, measurements of LIMA were performed from the parasternal area of the left second intercostal space in order to reduce the error margin.

In several studies , mean Vmax values of LIMA showed significant reduction in early and late postoperative measurements compared to control group [₅,₆]. In our study, Vmax values of LIMA decreased in the postoperative period compared to preoperative levels.

Since coronary arteries are filled during diastolic phase, increase in diastolic flow velocity was interpreted as adaptation to coronary circulation. Hence, diastolic flow samples give reliable information about graft function. Ichikawa et al. [₆] and Catalyurek et al. [₅] detected that the mean Vmin values of LIMA is increased in postoperative period. In our study, we observed a gradual increase in Vmin values of LIMA together with reduction in Vmin values of LVA, LSA and LBA. Catalyurek et al. [₅], emphasized that among the semiquantitative parameters, PI and RI values are decreased in the postoperative period compared to control group and preoperative levels. In our study, these 2 parameters were found to be in accordance with the literature.

Although Ichikawa et al. [₆] defined that the diameter of LIMA decreased compared to the control group during postoperative sixth months and became equalized at fifth and tenth year, Catalyurek et al. [₅] stated that the diameter decreased in early postoperative period and became equalized to the control group in the late postoperative period. Our results showed that the diameter of LIMA decreased in the early and late postoperative periods compared to the preoperative evaluation and and the diameter of the RIMA .

It was thought that no alteration in vessel diameters during the late postoperative controls may be a result of using β-blockers, but no statistical difference was detected in comparison of these 2 groups. Besides, despite reduction in diameter of LIMA in the postoperative period, no significant change in RIMA diameter caused us to think that using β-blockers in postoperative period does not alter LIMA flow pattern.

Nasu et al. [₁₃] emphasized that the degree of stenosis at the proximal portion of the anastomosed LAD affects the postoperative LIMA flow volume. If there's a low grade proximal stenosis on the LAD, a string phenomenon may occur on the LIMA-LAD anastomosis and a reduction in LIMA flow may happen. If one of the major intercostal branches remains at the LIMA preparation stage, steal phenomenon may occur [₁₃,₂₆,₂₇].

In our study group, steal or string phenomenons were not detected in any of the patients in the control color Doppler US examination. We considered that 70% or more stenosis on the LAD proximal to the anastomosis region and dissection of all intercostal arteries may have an important role.

Since the patency rate of IEA is better than venous grafts, it is frequently used as an arterial conduit [₁]. Its diameter and length should be determined by US examination in the preoperative period and should be excised with a separate incision from the abdomen [₁₁, ₂₈]. Collateral circulation among LIMA and IEA gains importance especially in patients that have chronic lower extremity ischemia [₉,₁₀,₁₂]. Because the distal end of LIMA is ligated during preparation stage, blood flow to IEA from subclavian artery is obstructed. If diameters of IMA, SEA and IEA are >3 mm and are tortious, they can be considered as important collateral sources to lower extremity [₁₀]. Hayashida et al. [₁₀] showed the importance of LIMA circulation for the lower extremity in two patients whose collateral system formed by inferior mesenteric artery and lumbar arteries are obstructed. Shimuzi et al. [₉] established the LIMA and IEA diameters as 3 mm in a patient with left iliac artery occlusion and showed the importance of LIMA collaterals, they used LIMA and IEA conduits in coronary bypass surgery together with femoro-femoral crossover bypass. These studies showed the importance of the IMA-IEA collateral circulation in lower extremity ischemia.

Since our patients did not have chronic ischemia of the lower extremities, we detected a significant reduction in Vmax, Vmin, and PI values of IEA in early and late postoperative controls after occluding distal end of LIMA, there was no change in RI and diameter.

This study shows that there are important changes in flow characteristics of neighbouring arteries in LIMA graft used patients. Among those; the most important one is the reduction in Vmax, Vmin, PI and diameter of LVA. As it is known, vertebral artery originates from subclavian artery and is an important source for cerebral circulation. In case of proximal subclavian artery stenosis, steal from cerebral circulation may arise and this may be so severe that it may require surgical correction. In patients who don't have proximal subclavian artery stenosis and LIMA was used as a graft, significant reduction in flow samples of LVA showed us that amount of blood flow to cerebral circulation is diminished. We think that detailed investigations should be performed to evaluate the neurologic state due to this reduction.

In our study; despite Vmax and Vmin values of LSA decreased in the postoperative period, we didn't detect an alteration in flow pattern of brachial artery, thus decrease in effort capacity of left arm wasn't observed. Vmax ratio of LSA/LIMA was 2.0 preoperatively, 2.7 for the first control, 2.1 for the third control. While Vmin value of LIMA is increasing, Vmin value of LSA was decreasing.

If there is a concomitant stenosis at the subclavian artery proximal to the IMA that is used as a conduit in coronary artery revascularization or if a stenosis develops in the postoperative period , coronary-subclavian artery steal syndrome may arise. If there is a 20 mmHg or more blood pressure gradient among upper extremities prior to coronary bypass operation, detailed examination of this stenosis is necessary [₈,₂₉]. Non-selective aortic arch investigation must be performed together with angiography in order to detect the stenosis at subclavian or brachiocephalic arteries since IMA graft may be used in patients with peripheral vascular disease. Varying degrees of proximal subclavian artery stenosis can be detected during coronary angiography in 0.5-1.1 % of patients whose subclavian artery was imaged [₃₀]. While percutaneous interventions may be sufficient for the stenosis established at subclavian artery proximal to LIMA in the postoperative period, sometimes additional surgical procedures are required. Hennen et al. [₈] stated that they corrected the graft dysfunction due to proximal stenosis of LIMA graft with percutaneous interventions.

RIMA measurements may be preferred to compare the changes in flow pattern of LIMA usually. Sungun et al. [₂₃] stated that there is an evident increase in diameter and avarage flow values of LIMA compared to RIMA. In our study, we detected a decrease in PI values and an increase in RI values of RIMA at preoperative and early postoperative periods, however we couldn't find any alteration in other parameters.

While significant changes in flow patterns of LIMA, LSA, LVA, IEA and RIMA were detected in the early and late postoperative controls with color Doppler US in LIMA graft patients, no significant change occured on LBA. Further investigations will clarify the effects of these changes in the postoperative period and course of the long term changes in flow patterns.

Correspondence to

Ass.Prof.Turan EGE Trakya University Medicine Faculty Department of Cardiovascular Surgery 22030 Edirne/ TURKEY Phone: + 90 284 235 06 65 Fax: + 90 284 235 06 65 e-mail: turanege@ttnet.net.tr