Both skeletonized and pedicled internal thoracic arteries supply adequate graft flow after coronary artery bypass grafting even during intense sympathoexcitation

Abstract

The internal thoracic artery (ITA) is harvested by either the pedicled or the skeletonized technique in coronary artery bypass grafting (CABG), with no clear advantage of one technique over the other. We compared graft flow between the pedicled and skeletonized ITA grafts while varying myocardial oxygen demand. CABG was performed to the left anterior descending artery in five anesthetized dogs using a pedicled ITA graft and the graft was subsequently skeletonized. Graft flow was measured during stepwise electrical stimulation of the stellate ganglion. The baseline graft flow before sympathetic stimulation was higher in skeletonized (27.8 ± 1.9 ml/min) than that in pedicled ITA grafts (22.6 ± 2.7 ml/min) (P < 0.05). In both ITA grafts, however, graft flow increased to a similar level during sympathetic stimulation that doubled the double product, correlating with the double product. Based on these results, we conclude that metabolic demand can override the potential difference in sympathetic vasoconstriction in both pedicled and skeletonized ITA grafts.

Introduction

The internal thoracic artery (ITA) is the gold standard conduit for coronary artery bypass grafting (CABG) because of its long-term patency [1]. The ITA is harvested by either the pedicled or the skeletonized technique, and which of these two techniques is the better option has been the subject of an extended debate—with as yet no clear conclusion being drawn. Although some human studies [2–4] have demonstrated higher free (pre-anastomosis) flow through skeletonized grafts (with or without topical papaverine), suggesting that the loss of sympathetic nerve-mediated graft vasoconstriction confers an advantage, perfusion pressure was not controlled in these studies. In one study [5] in which the perfusion pressure was controlled, free flow even tended to be lower in skeletonized grafts prior to the administration of intravenous papaverine. Onorati et al. [6] found that graft flows were comparable between the two techniques in the absence of intraluminal papaverine, while Takami and Ina [7], in a comparison of the flow through the anastomosed graft, found that flow was higher through the skeletonized graft.

Flow in the anastomosed graft is likely to be largely dependent on myocardial oxygen demand, suggesting the importance of comparing the flow between the pedicled and skeletonized ITA grafts under varying conditions of myocardial oxygen demand. If the skeletonization procedure were to result in an increased flow capacity, surgeons may be able to perform additional anastomoses to other vessels using the skeletonized ITA, thereby making the skeletonized ITA procedure even more advantageous. If the skeletonization procedure were not able to increase flow capacity, the skeletonized ITA would not be recommended for additional use due to a higher flow reserve. We hypothesized that the skeletonized ITA would have larger flow capacity due to the loss of sympathetic nerve-mediated graft vasoconstriction.

Materials and methods

Animal preparation

Animal care was provided in accordance with the Guiding Principles for the Care and Use of Animals in the Field of Physiological Sciences approved by the Physiological Society of Japan. All protocols were approved by the Animal Subject Committee of the National Cerebral and Cardiovascular Center. Five adult mongrel dogs (weighing 24–35 kg) were anesthetized with intravenous pentobarbital sodium (25 mg/kg) and intubated endotracheally for artificial ventilation with isoflurane and 100% O₂. After a median sternotomy, the heart was suspended in a pericardial cradle. To measure systemic arterial pressure, we placed a fluid-filled catheter in the left subclavian artery via the left brachial artery and connected it to a pressure transducer (DX-200; Nihon Kohden, Tokyo, Japan). The junction of the inferior vena cava and the right atrium was taken as the reference point for zero pressure. An ultrasonic flowmeter (20A594; Transonic Systems, Itaca, NY) was placed around the ascending aorta to measure cardiac output. Electrocardiography leads were also placed for the monitoring electrocardiogram. A catheter was inserted into the femoral vein for fluid replacement (1 ml/kg/h of Ringer’s solution). All protocols were performed under open chest conditions.

Pedicled ITA grafting

The left internal thoracic artery (LITA), together with the surrounding veins, muscle, and fascia, was harvested as a pedicled graft using electrocautery. The LITA was harvested from the bifurcation of the musculo-phrenic and superior epigastric arteries up to the upper margin of the first rib or higher. All intercostal branches of the LITA were ligated. After systemic heparinization, the LITA was clamped, and the distal end of the LITA was cut and anastomosed to the left anterior descending artery (LAD). The same surgeon (D.U.) performed the LITA–LAD anastomosis without cardiopulmonary bypass. The heart and the LAD were stabilized using a compression-type mechanical stabilizer (Mini-CABG system; United States Surgical Corporation, Norwalk, CT). A shunt tube was inserted into the LAD to prevent myocardial ischemia during anastomosis. The anastomosis was placed in the mid-LAD [8]. The anastomosis was created using a continuous 7-0 polypropylene suture. The proximal LAD was first ligated after the LITA–LAD anastomosis, and then the LITA was declamped. An angiography was performed after the anastomosis to confirm the absence of stenosis or spasm in the LITA–LAD anastomosis. The LITA graft was sprayed with dilute papaverine (4 mg/ml) to prevent spasm. An ultrasonic flowmeter (2.5S261; Transonic Systems) was placed around the LITA just proximal to the anastomosis. The left stellate ganglion was carefully exposed through a median sternotomy, and a pair of platinum electrodes was attached to it without decentralization. The nerve and electrodes were covered with a mixture of silicone gel (Kwik-Sil; World Precision Instrument, Sarasota, FL). Protocol 1, described below, was carried out following the pedicled LITA grafting.

Skeletonized ITA grafting

Following the completion of protocol 1, the tissue surrounding the graft (including fascia and lymphatics) was stripped up to the most proximal part of the LITA graft in order to skeletonize the LITA graft. The side branches of the LITA were ligated. Fat tissue around the graft was removed as completely as possible based on macroscopic inspection. The adventitia was left as the outermost layer of the graft. The graft was not touched directly with forceps. The graft was sprayed with dilute papaverine (4 mg/ml). After skeletonizing the LITA graft, protocol 2 followed.

Experimental protocols

Since skeletonization always followed pedicled harvesting, protocol 1 (pedicled LITA graft flow measurement) was performed before protocol 2 (skeletonized LITA graft flow measurement) in all dogs. The stimulation of the left sympathetic stellate ganglion for adjusting the voltage amplitude was performed at least 30 min before protocol 1 was initiated.

Protocol 1

The left sympathetic stellate ganglion was electrically stimulated at least 30 min after the completion of the experimental preparation of the pedicled LITA grafts. The frequency of stimulation was increased stepwise from 0 to 10 Hz with increments of 2 Hz. Each step was maintained for 60 s. The pulse duration of the stimulus was set at 5 ms. The voltage amplitude of stimulation (2–5 V) was adjusted in each animal to yield an increase in arterial pressure of approximately 30 mmHg with 10 Hz stimulation. Graft flow, arterial pressure, and cardiac output were recorded for 7 min, which included a 2-min baseline and 5 min of stimulation. These data were sampled at 200 Hz using a 12-bit analog-to-digital converter [AD12-16U(PCI)E; CONTEC, Osaka, Japan] and stored on the hard disk of a dedicated laboratory computer system.

Protocol 2

At least 30 min after the completion of the experimental preparation of the skeletonized LITA grafts, the left sympathetic stellate ganglion was electrically stimulated in a similar fashion to protocol 1, while all variables were recorded and stored.

Data analysis

Heart rate was calculated from the arterial pressure waveform. Myocardial oxygen demand was estimated as double product (pressure–rate product) and calculated as the product of systolic arterial pressure and heart rate [9]. All variables were averaged during the last 20 s of each electrical stimulation level.

Statistical analysis

All data are presented as the mean ± standard error (SE). In each protocol, one-way repeated measures analysis of variance (ANOVA) followed by Dunnett’s test was used to compare variables at each stimulation against the baseline value. The paired t test was used to compare variables between pedicled and skeletonized LITA grafts at each stimulation level. Linear regression analysis was used to examine the relationship between the double product and graft flow. Differences were considered to be significant at a threshold of P < 0.05.

Results

Prior to sympathetic stimulation, baseline graft flow (under spontaneous sympathetic outflow) was greater in skeletonized ITA than pedicled ITA (Table 1). Other hemodynamic parameters, including heart rate, cardiac output, mean arterial pressure, and double product, did not differ significantly regardless of harvesting technique.

Table 1 Hemodynamic parameters and graft flow before stimulation

Graft flow patterns at baseline and under sympathetic stimulation are shown in Fig. 1a. Sympathetic stimulation increased graft flow (P < 0.05) similarly in skeletonized and pedicled ITA grafts, and maximal flow was comparable to each other at 10-Hz stimulation [nonsignificant (NS) difference] (Fig. 1b). Increases in systemic arterial pressure and heart rate did not differ significantly between the two techniques (Fig. 2), and increases in myocardial oxygen demand in response to sympathetic stimulation, as estimated by double product, were likewise similar.

Fig. 1

a Typical representative recording of graft flow with pedicled and skeletonized internal thoracic arteries (ITAs) during sympathetic nerve stimulation. b Mean graft flow with pedicled (closed circle) and skeletonized (open circle) ITAs during sympathetic nerve stimulation. Data are shown as the mean ± standard error (SE). ^†P < 0.05 vs. baseline, ^‡P < 0.01 vs. baseline, *P < 0.05 pedicled vs. skeletonized

Fig. 2

Changes in mean arterial pressure, cardiac output, heart rate, and double product with pedicled (closed circle) and skeletonized (open circle) ITAs during sympathetic nerve stimulation. Data are shown as the mean ± SE. ^† P < 0.05 vs. baseline, ^‡ P < 0.01 vs. baseline

Graft flow (y) correlated well with the double product (x) in both pedicled (y = 2.6 × 10⁻³ x − 8.4, R ² = 0.73) and skeletonized ITA (y = 2.3 × 10⁻³ x − 0.7, R ² = 0.69). The slope and y-intercept did not differ statistically between the two techniques (Fig. 3).

Fig. 3

Scatter plots and regressions between the double product and graft flow with pedicled (closed circle, solid line) and skeletonized (open circle, dashed line) ITAs. Regression lines did not differ between the two groups. y Graft flow, x double product

Discussion

The choice of either skeletonized or pedicled ITA grafts for CABG may be an important decision from both the technical and clinical viewpoints; however, clear evidence demonstrating the advantage of either method over the other is not yet available. In this study, we have shown that graft flow increased to a similar level during maximal sympathetic stimulation in both pedicled and skeletonized ITA grafts. These results do not support our hypothesis that the skeletonized ITA would provide larger flow capacity and indicate that coronary vasodilatation in response to increased myocardial oxygen demand is a stronger determinant of graft flow than any possible increase in the vascular resistance of ITA itself. Our study also demonstrates that both skeletonized and pedicled ITAs were able to supply adequate graft flow after CABG even during intense sympathoexcitation.

There are several possible explanations for the difference in graft flow under baseline conditions. First, a loss of sympathetic innervation in the skeletonized graft may have dilated the ITA relative to the pedicled graft under baseline conditions. In support of this explanation, Takami et al. [7] reported that the diameter of the ITA just proximal to the anastomosis is significantly larger in the skeletonized ITA than that in the pedicled ITA. Dönmez et al. [10] reported that the diameter of ITA becomes statically larger by the stellate ganglion blockade. In a preliminary study, we observed that electrical stimulation of the stellate ganglion decreased ITA flow before harvest. Therefore, vasoconstriction may occur in the pedicled ITA during sympathetic stimulation. However, in this study we did not perform simultaneous measurements of the graft flow and diameter because the use of contrast medium in angiography may have affected the graft flow through its vasodilatative effect on the coronary artery [11].

Another explanation may be the difference in background sympathetic tone. As the skeletonized graft flow was always studied in the later phase of the experiment, when background sympathetic tone and myocardial metabolic demand may be higher, skeletonized graft flow may have been higher for this reason. The presence of similar hemodynamics during the two protocols, however, does not directly support this explanation. The hemodilution seen predominantly in the later phase of the experiment may also have contributed to higher flow through the skeletonized graft under baseline conditions.

The fact that graft flows were similar between the skeletonized and pedicled ITAs during maximal sympathetic excitation indicates that the resistance of the ITA graft was much smaller than that of the native coronary bed, even when the coronary bed was maximally dilated to meet the oxygen demand present with maximal sympathetic stimulation. In other words, both the skeletonized and pedicled ITAs would appear to provide sufficient flow reserve to the LAD area. In contrast, it has been reported that free flow, which may represent the maximal flow capacity of the ITA itself, is greater in the skeletonized ITA than in the pedicled ITA [2, 3]. Despite these previous findings, because the maximally dilated native coronary bed would be the most practical downstream conduit to test the difference between the skeletonized and pedicled ITAs, we believe that the difference in sympathetic innervation does not affect the maximal flow significantly under anastomosed conditions.

In addition to the effects of downstream resistance, local mechanisms would also contribute to the observed difference in flow between the pedicled and skeletonized ITA grafts. Complete sympathetic denervation with the local application of phenol to the skeletonized ITA further increased graft flow (unpublished observation), suggesting that there remains a certain sympathetic innervation in the skeletonized ITA. Even though sympathetic denervation may not be complete after skeletonization, we believe that our skeletonization did not differ greatly from those clinically performed by surgeons. Deja et al. [4] reported that skeletonization increases the reactivity of ITA to norepinephrine in vitro. Their study may support our results. Prior to sympathetic stimulation but under spontaneous sympathetic outflow, the amount of endogenous norepinephrine release to the skeletonized ITA may be relatively smaller than that to the pedicled ITA; as such, the sympathetic vasoconstriction would be negligible in the skeletonized ITA. This may explain the larger graft flow in the skeletonized ITA prior to sympathetic stimulation. Under maximal sympathetic stimulation, however, hyperreactivity to endogenous norepinephrine in the skeletonized ITA may cause the sympathetic vasoconstriction similar to that occurring in the pedicled ITA. This local mechanism may also partly account for why graft flow was comparable between the pedicled and skeletonized ITAs during maximal sympathetic stimulation. Although the results from several pharmacological studies suggest that norepinephrine-induced vasoconstriction does occur in the ITA [12, 13], there have been no reports assessing the tissue norepinephrine concentration of ITA during sympathetic stimulation. Further investigations are necessary to gain an understanding of the difference in norepinephrine reactivity between the pedicled and skeletonized ITAs.

Some publications have reported several advantages of the skeletonized ITA grafts other than the potential increase in graft flow at rest [1, 14]. Firstly, skeletonization lengthens the ITA, thereby providing access to more distal targets in the coronary artery [15]. Second, skeletonization improves blood supply to the sternum (measured by single photon emission computed tomography) [16] compared with pedicled harvesting. Third, skeletonization decreases the incidence of postoperative respiratory dysfunction because of less invasive harvesting (i.e., preserved pleural integrity in skeletonized ITA vs. pleurotomy in pedicled ITA) [17, 18]. Lastly, skeletonization markedly reduces anterior chest pain and dysesthesia 3 months after surgery [19]. In contrast to these advantages, skeletonization has the minor disadvantages of greater technical difficulty, longer harvesting duration, and potential damage to the graft.

Limitations

This study has several limitations. First, because skeletonized graft flow measurements always follow pedicled graft flow measurements in the same dog, the effect of time sequence on graft flows cannot be ruled out. Nevertheless, the similar hemodynamic response to sympathetic stimulation between protocol 1 and 2 (Fig. 2) suggests that the animal conditions did not deteriorate considerably. Second, the perfusion area of LITA was limited to the LAD region. If we had used a much larger perfusion area of LITA, the possible small difference between the pedicled and skeletonized ITAs may have been revealed. Third, a histological comparison between the pedicled and skeletonized ITA was not performed because the pedicled ITA was always skeletonized after the protocol 1, and the tissue samples from the pedicled ITA could not be obtained. Further investigations that include histological comparison are necessary for examining the effect of skeletonization on sympathetic innervations.

Conclusion

Both the pedicled and skeletonized ITA techniques supplied similar, adequate blood flow to the LAD, meeting myocardial oxygen demand during sympathetic excitation. Metabolic demand can override the possible difference in sympathetic vasoconstriction, increasing the flow in both pedicled and skeletonized ITA grafts to a similar extent when they are anastomosed to LAD. The results of this study have an important implication in terms of clinical application. Following anastomosis, graft flow is highly variable and is dependent on myocardial oxygen demand. Because the quality of CABG may be judged based on flow through anastomosed grafts, one has to take into consideration the potential change in flow in response to myocardial oxygen demand.