Author + information
- Received February 12, 2007
- Revision received September 17, 2007
- Accepted September 17, 2007
- Published online January 15, 2008.
- David Glineur, MD⁎ (, )
- William D’hoore, MD, PhD,
- Gebrine El Khoury, MD,
- Sixte Sondji, MD,
- Gregory Kalscheuer, MD,
- Jean-Christophe Funken, MD,
- Jean Rubay, MD, PhD,
- Alain Poncelet, MD,
- Parla Astarci, MD,
- Robert Verhelst, MD,
- Philippe Noirhomme, MD and
- Claude Hanet, MD, PhD
- ↵⁎Reprint requests and correspondence:
Dr. David Glineur, Service de Chirurgie Cardiovasculaire et Thoracique, Cliniques Universitaires Saint-Luc–U.C.L.90, Avenue Hippocrate 10/6107, 1200 Brussels, Belgium.
Objectives The purpose of this study was to define the pre-operative angiographic variables that could influence graft patency and flow pattern.
Background Saphenous vein grafts (SVG) and pedicled right gastroepiploic artery (RGEA) grafts are routinely used to revascularize the right coronary artery (RCA). Little is known about the predictive value of objective pre-operative angiographic parameters on the 6-month graft patency and on the interest of these parameters to select the optimal graft material in individual cases.
Methods We prospectively enrolled 172 consecutive patient candidates for coronary revascularization. Revascularization of the RCA was randomly performed with SVG in 82 patients or with the RGEA in 90 patients. Both groups were comparable with respect to all pre-operative continuous and discrete variable and risk factors. All patients underwent a systematic angiographic control 6 months after surgery. Pre-operative angiographic parameters included minimal lumen diameter (MLD), percent stenosis and reference diameter of the RCA measured by quantitative angiography (CAAS II system, Pie Medical, Maastricht, the Netherlands), location of the stenosis, run off of the RCA, and regional wall motion of the revascularized territory.
Results A significant difference in the distribution of flow patterns was observed between SVG and RGEA. In multivariate analysis, graft-dependent flow pattern was significantly associated with both MLD and percent stenosis of the RCA in the RGEA group but with percent stenosis only in the SVG group. In the RGEA group, the proportion of patent grafts was higher when MLD was below a threshold value lying in the third MLD quartile (0.77 to 1.40 mm).
Conclusions Pre-operative angiography predicts graft patency in RGEA, whereas the flow pattern in SVG is significantly less influenced by quantitative angiographic parameters.
The choice of the ideal bypass conduit for each diseased coronary vessel remains a subject of intense controversy. For the left coronary system, bilateral internal thoracic arteries have clearly demonstrated their superiority over all other types of grafts (either venous or arterial) in terms of patency, freedom from arteriosclerosis, and survival benefit (1,2). Conduits considered acceptable to revascularize the right coronary artery (RCA) system include the right gastroepiploic artery (RGEA) (3), the right internal thoracic artery (RITA) in situ or used in a Y graft configuration (4), the free radial artery (RA) implanted into the aorta or into the left internal thoracic artery (5), and the saphenous vein graft (SVG) (6).
The influence of the type of conduit implanted to the RCA system on clinical results remains unclear, and the complementary conduit of choice for this system has yet to be determined. No superior patency rate for any of these grafts to the RCA has been clearly established (7). The patency results of the RITA on the RCA have been reported lower than those obtained when used for the left coronary system (8). One evaluation of the SVG to the RCA territory revealed surprisingly good clinical and angiographic results after long-term follow-up (6). The patency rates of RA and RGEA are highly dependent on the degree of stenosis of the native vessel, and their use remains limited because of its association with a high risk of graft failure, owing to competitive flow (9,10). The flow capacity of the RGEA under maximal stress conditions has also been questioned (11).
Because of these recent studies, debating the potential benefit of an arterial graft to revascularize the RCA, we prospectively randomized 2 types of graft, the RGEA and the SVG, which are routinely used in our department, and analyzed the predictive value of objective pre-operative coronary angiographic parameters on 6-month graft patency.
From April 2003 to July 2006, 1,297 consecutive patients underwent isolated bypass surgery in our institution. Of this group, 360 patients met the entrance criteria for randomization (Table 1) and 236 patients were actually randomized. Among those 236 patients, 172 had a pre-operative angiogram suitable for biplane quantitative angiographic analysis and are the subject of the present study. Right coronary artery grafting was performed with an RGEA in 90 patients and with an SVG in 82. Four patients initially allocated to the SVG group were excluded for protocol deviation; in these cases, the surgeon decided to deviate from the assigned revascularization strategy in favor of the RGEA group (Fig. 1). Patients’ basal demographic and clinical characteristics are shown in Table 2. The RGEA and the SVG were harvested and grafted as previously described (12,13). The RGEA diameter was measured pre-operatively with a 1.5-mm probe. The RCA was systematically grafted distally to the bifurcation either on the posterior descending artery (PDA) or on the posterolateral branch (PLA) in order to revascularize the RCA distal territory with a single graft. All patients underwent a systematic angiographic control between the 6 to 8 months after coronary artery bypass grafting. The study protocol was approved by the ethics committee of our institution. All patients gave informed consent at the time of bypass surgery and before the angiographic investigations.
Pre-operative angiograms were analyzed with the CAAS II system (Pie Medical, Maastricht, the Netherlands) as previously described (14). The minimal lumen diameter (MLD), percent diameter stenosis (%), and reference diameter of the native RCA were measured in 2 near orthogonal projections, and the mean value of these measurements was considered for statistical analyses.
All pre- and post-operative angiograms were independently reviewed by 2 investigators and discrepancies in patency scores were reviewed by a third investigator and resolved by consensus. All investigators were blinded to the experimental conditions. Graft patency is scored as 0 for an occluded graft, 1 when the flow from the native coronary artery is dominant, 2 when flow supply from the native coronary and from the graft is balanced, 3 when the native coronary is fully opacified by the graft, and 4 when the native coronary is fully opacified by the graft only (occluded or suboccluded coronary native vessel). A graft was considered “not functional” for patency scores of 0 to 2 and “functional” for patency scores of 3 or 4. The regional wall motion of the inferior wall was defined as normal, hypokinetic, or akinetic on the pre-operative angiogram or echocardiogram. The run off was graded according to the extent of the coronary territory beyond the stenosis on the diagnostic angiogram: incomplete when only 1 branch, either posterior descending or posterolateral and complete when the postero-descending and the posterolateral territory were subject to revascularization.
Data are expressed as mean ± 1 standard deviation. To compare the pre-operative characteristics of experimental and control groups, chi-square and t tests were used for categorical and continuous variables, respectively (Table 2). To study the relationships between MLD, percent stenosis, and graft function, variables MLD and percent stenosis were transformed into 4° ordinal variables on the basis of the approximate quartiles of their respective distributions. Recoding was justified here by the extreme skewness of data distribution. For MLD, boundaries were 0 mm (first quartile), 0.77 mm (median), and 1.40 mm (third quartile), resulting in a 4° MLD with values 0, 1, 2, and 3 corresponding to 0, 0.01 to 0.76, 0.77 to 1.40, and >1.40 mm, respectively. For percent stenosis, boundaries were 48% (first quartile), 65% (median), and 100% (third quartile), resulting in a 4° percent stenosis with values 0 = 0% to <48%, 1 = 48% to <65%, 2 = 65% to <100%, and 3 = 100%.
Mantel-Haenszel chi-square test was used to test the associations between graft function and type of graft, 4° MLD, and 4° percent stenosis. Multiple logistic regression modeling was then performed to obtain a parsimonious, clinically sound model.
All p values are 2-tailed. A p value <0.05 was considered significant. SAS software (version 9.1, SAS Institute Inc., Cary, North Carolina) was used to perform the statistical analysis.
Pre-operative characteristics of the RCA lesion are shown in Table 2. At the 6-month angiographic follow-up, the difference between the proportions of functional RGEA (62%) and SVG (85%) was statistically significant (odds ratio 3.54; 95% confidence interval 1.68 to 7.47). We hypothesized that the differences were partly explained by the effects of pre-operative RCA characteristics. The proportion of functional RGEA grafts was strongly dependent on the severity of the RCA stenosis (Table 3), only 5% of the RGEA being functional when the MLD of the RCA was in the highest quartile (>1.40 mm). Conversely, only a weak association between MLD or percent stenosis and the function of SVG was observed, indicating the presence of an interaction between either MLD or percent stenosis and the type of graft. Figure 2 displays the interaction between MLD and type of graft.
In the multiple logistic regression modeling strategy, interaction terms were then introduced under the form of cross-products (MLD × type of graft, and percent stenosis × type of graft). The following variables were entered: type of graft (RGEA vs. SVG), MLD, percent stenosis, and the 2 interaction terms. Results are shown in Table 4. The MLD, percent stenosis, and both interaction terms were significantly associated with graft function. No association was observed between graft patency or function and the regional wall motion of the RCA territory or the run off of the RCA graft. The strongest association was observed for the interaction between MLD and type of graft. Separate logistic regression models for SVG and RGEA are presented in Table 5. The MLD and percent stenosis, when tested together in this model, were not significant predictors of graft function for the SVG, whereas for RGEA they were significantly associated with the graft function. When testing the significance of MLD or percent stenosis alone in SVG grafts, only percent stenosis was significant (beta estimate = 0.64, SE = 0.28, p = 0.0250).
When the population was divided in 4 equivalent groups on the basis of the severity of the RCA stenosis assessed by MLD and by percent stenosis, the mean patency score of RGEA decreases abruptly in the approximate quartiles corresponding to the less severely obstructive RCA disease. In contrast, the influence of MLD and of percent stenosis of the RCA on the patency score of SVG is significantly less pronounced and completely disappears when MLD is adjusted to percent stenosis. The MLD values approximately 0.77 to 1.4 mm and percent stenosis approximately 48% to 64% seemed to discriminate between functional and nonfunctional RGEA. In contrast, no threshold values could be established for the SVG, the proportion of functional grafts remaining ≥70% in the higher MLD approximate quartiles.
This study shows that in patients randomly treated with either an RGEA or an SVG for a stenosis of the RCA, the proportion of grafts found occluded on systematic 6-month angiographic control is low and independent of the type of graft conduit; however, the flow pattern is significantly more affected by the angiographic characteristics of the grafted RCA for RGEA than for SVG. Consequently, most SVG but only a minority of RGEA remained functional when the MLD of the RCA was larger than a threshold value close to 1.0 mm.
Previous angiographic follow-up studies of RGEA have suggested that the graft patency might be affected by the stenosis of the recipient RCA, most likely as a result of competitive flow (3,15,16). The evaluation of stenosis severity was based on visual interpretation of post-operative coronary angiogram at the time of angiographic follow-up. In the present study, the evaluation of RCA stenosis severity was done on pre-operative angiogram to eliminate the possible influence of coronary remodeling secondary to changes in native coronary flow pattern after surgery. In addition, this evaluation was performed with a validated quantitative angiographic analysis system in order to eliminate subjectivity in the assessment of stenosis severity. Although quantitative angiography has been extensively used in interventional cardiology for many years for precise indications and to appreciate the results of percutaneous interventions, little is known about the influence of quantitative angiographic parameters of coronary vessels on the outcome of coronary bypass grafts.
Several factors could contribute to the greater predictive value of RCA stenosis severity on the flow pattern of RGEA. Compared with an SVG, which is directly anastomosed to the ascending aorta and can be considered as a first branch off of the aorta, the pedicled RGEA is the fourth branch originating from a visceral system. This more distal origin from a complex branching arterial system, together with a lumen diameter usually smaller than that of venous grafts, might contribute to a lower flow capacity of RGEA compared with SVG. Intraoperative measurements of pressure waveforms at the tip of RGEA have shown that mean diastolic pressure is significantly lower than that in the ascending aorta or at the tip of a mammary artery graft (17). In addition, the gastroepiploic artery is prone to vasoconstriction and might become more resistive in responses to various pharmacologic and physiologic stimuli (18). In contrast, the range of vasomotor response of SVG is smaller than that of grafted coronary arteries (14). In the presence of an RCA stenosis of moderate severity, the pressure drop across the stenosis might be small, particularly at rest, when myocardial blood flow demand is minimal. We can hypothesize that the perfusion of the distal coronary bed is mainly provided by the less resistive conduit, which explains the superiority of SVG over RGEA in these conditions.
Quantitative angiographic analysis of pre-operative angiograms provides information likely to influence the choice of the most adequate graft conduit in patients who are candidates for coronary bypass grafting of the RCA. In the present study, an MLD above the third quartile (0.77 to 1.4 mm) or a stenosis <55% predicted an unfavorable flow pattern in RGEA but not in SVG at 6-month follow-up. In these stenoses of intermediate severity, the SVG should be preferred to the RGEA.
This study was supported by grant no. 3.4600.04 from the Fonds de la Recherche Scientifique Médicale, Brussels, Belgium.
- Abbreviations and Acronyms
- minimum lumen diameter
- posterior descending artery
- posterior lateral artery
- radial artery
- right coronary artery
- right gastroepiploic artery
- right internal thoracic artery
- saphenous vein graft
- Received February 12, 2007.
- Revision received September 17, 2007.
- Accepted September 17, 2007.
- American College of Cardiology Foundation