Skip to main content
Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease logoLink to Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease
. 2026 Sep 9;15(19):e051333. doi: 10.1161/JAHA.126.051333

Impact of Left Internal Mammary Artery–to–Left Anterior Descending Artery Anastomotic Location on Graft Patency: A Secondary Analysis of the PATENCY Trial Data Set

Chunyuan Wang 1,#, Meice Tian 1,#, Yang Wang 2, Zhihui Hou 3, Wei Feng 1, Yan Zhang 1,✉, Zhan Hu 1,✉
PMCID: PMC13745009  PMID: 42757942

Abstract

Background

The impact of the left internal mammary artery (LIMA)–to–left anterior descending artery (LAD) anastomotic location on graft patency remains uncertain.

Methods

This study included 1597 patients with isolated proximal LAD lesions from the PATENCY (Graft Patency Between no‐touch Vein Harvesting Technique and Conventional Approach in Coronary Artery Bypass Graft Surgery) trial who underwent coronary artery bypass grafting. The relative anastomotic site (RAS) of the LIMA–LAD anastomosis was quantified using postoperative coronary computed tomography angiography and defined as the ratio of anastomotic height/heart height. The primary end point was LIMA occlusion at 3 years, and the secondary end point was major adverse cardiac events. Multivariable logistic regression was used to evaluate the association between RAS and 3‐year LIMA occlusion, and Cox proportional hazards regression was used to evaluate associations with time‐to‐event clinical outcomes.

Results

The exploratory RAS threshold was 46%. Distal anastomosis (RAS >46%) was associated with a significantly lower 3‐year LIMA occlusion rate compared with proximal anastomosis (RAS ≤46%) (2.3% versus 6.3%; adjusted odds ratio, 0.37 [95% CI, 0.20–0.68]; P=0.002). At 3 years, major adverse cardiac event rates were similar between groups (3.4% versus 4.4%; adjusted hazard ratio, 0.68 [95% CI, 0.40–1.15]; P=0.16).

Conclusions

In patients with isolated proximal LAD disease, a more distal LIMA–LAD anastomotic location may be associated with a lower risk of graft occlusion at 3 years. Further prospective validation is warranted.

Keywords: anastomotic location, coronary artery bypass grafting, graft patency, left internal mammary artery

Subject Categories: Revascularization


Nonstandard Abbreviations and Acronyms

LIMA

left internal mammary artery

MACE

major adverse cardiac event

RAS

relative anastomotic site

Clinical Perspective.

What Is New?

  • The location of the left internal mammary artery (LIMA)–to–left anterior descending artery (LAD) anastomosis is associated with graft patency, with a more distal anastomotic position linked to a lower rate of LIMA occlusion compared with a proximal position (2.3% versus 6.3%; P=0.002).

What Are the Clinical Implications?

  • For patients with isolated proximal LAD stenosis, a mid‐to‐distal LIMA–LAD anastomotic location provides graft patency noninferior to proximal anastomosis, supporting flexibility in surgical planning rather than prioritizing a proximal target.

  • Prospective trials are warranted to further compare LIMA patency and patient outcomes based on LIMA‐LAD anastomotic location.

Left internal mammary artery (LIMA)–to–left anterior descending artery (LAD) grafting is a cornerstone of coronary artery bypass grafting (CABG) surgery, offering improved outcomes for patients undergoing the procedure. 1 , 2 LIMA graft failure following CABG is strongly linked to adverse patient events, 3 , 4 underscoring the importance of maximizing LIMA patency to enhance long‐term outcomes.

The location of the LIMA‐LAD anastomosis varies significantly among different patients or surgeons, which can be located at the proximal, middle, or distal segment of the LAD. Different locations present distinct technical challenges, anatomic variations, and hemodynamic conditions, all of which could potentially influence graft patency. Despite this variability, the relationship between anastomotic site and LIMA patency remains unexamined, and there is no consensus on the optimal location for achieving superior outcomes.

Based on the robust imaging follow‐up data from the PATENCY trial, 5 this study seeks to systematically and quantitatively evaluate the potential impact of anastomotic location on LIMA graft patency.

METHODS

The data underlying this article cannot be shared publicly because of the policy of data sharing of our center. The data will be shared on reasonable request to the corresponding author.

Population

This study is a retrospective secondary analysis of the PATENCY trial data set. The design and results of the PATENCY trial have been previously published. 5 , 6 In brief, the PATENCY trial was a multicenter, randomized trial designed to compare the 3‐ and 12‐month patency between saphenous vein grafts harvested by conventional or no‐touch techniques. The protocol of the trial was approved by the institutional review board (2016‐827). All participants provided written informed consent to participate in the trial.

The present study focuses on patients with isolated proximal stenosis of LAD who underwent LIMA‐to‐LAD grafting with only 1 anastomosis on LAD. A total of 2638 patients enrolled in the PATENCY trial from 7 centers eventually received CABG surgery. A total of 202 patients who received non–LIMA‐to‐LAD grafting, multiple LAD anastomoses, and/or LAD endarterectomy were excluded. The preoperative angiography data of the remaining 2436 patients were reviewed. A total of 451 patients with significant stenoses (>70%) in the mid‐to‐distal segment of LAD and 65 patients with missing preoperative angiograms were excluded. A total of 323 patients with unavailable postoperative coronary computed tomography angiography (CCTA) images were also excluded, because the site of LIMA‐to‐LAD anastomosis in these patients could not be defined. Finally, 1597 patients were included in the present study (Figure 1).

Figure 1. Study flowchart.

Figure 1

CABG indicates coronary artery bypass grafting; CCTA, coronary computed tomography angiography; LAD, left anterior descending artery; and LIMA, left internal mammary artery.

Coronary Artery Bypass Grafting

CABG surgeries were performed by experienced surgeons who had completed at least 100 cases. The LIMA was harvested using either skeletonized technique for patients receiving bilateral internal mammary artery grafting (1.4%) or pedicled technique for patients with single LIMA grafting (98.6%). The use of cardiopulmonary bypass was left to the individual surgeon's discretion.

Definition of the Location of LAD Stenosis

Preoperative angiography was used to determine the location of stenosis in LAD. The default angiographic view for interpretation was set to the right anterior oblique view. In this view, the LAD was equally divided into 3 segments, and the degree of stenosis in each segment was recorded. Patients with at least 1 significant stenosis (>70%) in the mid or distal segment of LAD were excluded. For these patients, performing a single anastomosis either proximal or distal to the mid‐to‐distal lesions may not achieve complete revascularization of the LAD territory.

Definition of the Relative Anastomotic Site of LIMA‐LAD Anastomosis

The postoperative CCTA was used to quantify the relative anastomotic site (RAS) of LIMA‐to‐LAD grafting. Figure 2 illustrates the calculation process of RAS. First, the following 3 levels were determined on CCTA images: (1) the top plane of the pulmonary artery sinus; (2) the plane where LIMA and LAD converge; and (3) the bottom plane of the apex of the heart. Then, the height of the anastomosis (distance between the top plane of the pulmonary sinus and the plane of the anastomosis) and the height of the heart (distance between the top plane of the pulmonary sinus and the bottom plane of the apex of the heart) were calculated. The RAS was defined as the ratio of the anastomosis height/the heart height.

Figure 2. Definition of RAS of LIMA‐LAD grafting.

Figure 2

LAD indicates left anterior descending artery; LIMA, left internal mammary artery; and RAS, relative anastomotic site.

Validation of the RAS Measurement

To evaluate measurement reproducibility of the RAS metric, a random sample of 100 patients was selected. RAS measurements were independently performed by 2 observers blinded to clinical outcomes. To assess intraobserver reproducibility, 1 observer repeated all measurements after an interval of 1 week. Interobserver and intraobserver agreement were assessed using intraclass correlation coefficients.

To evaluate the anatomic validity of the RAS metric, an additional validation study was performed in a random sample of 300 patients (150 proximal RAS and 150 distal RAS). Using 3‐dimensional CCTA reconstruction, the total LAD length and the actual distance from the LAD origin to the LIMA‐LAD anastomosis were directly measured (Figure S1).

The actual relative anastomotic position was calculated as: actual anastomotic position/total LAD length. Differences between groups were compared to determine whether the RAS classification corresponded to the true anatomic location of the anastomosis along the LAD.

Study End Points

The primary end point was LIMA occlusion at 3 years postoperatively, assessed by CCTA follow‐up. Patients enrolled in the PATENCY trial were scheduled to receive CCTA follow‐up at 3 months, 1 year, and 3 years after CABG surgery. According to the FitzGibbon criteria, 7 the occlusion of LIMA was defined as the total absence of contrast in any segment of this conduit. The results of CCTA were independently reviewed by 2 radiologists at a central core laboratory.

The secondary end point was time‐related major adverse cardiac events (MACEs) from the time of CABG, defined as the composite of death from any cause, myocardial infarction, or repeat revascularization.

Statistical Analysis

Continuous variables are presented as either mean±SD or median and interquartile range and compared using the Student t‐test or Mann–Whitney U test according to their distributions. Categorical variables are presented as count and percentage and compared using the χ2 test or Fisher exact test as appropriate.

For the end point of 3‐year LIMA occlusion, multivariable logistic regression models were constructed to evaluate the association between RAS and graft patency. Covariates were selected based on clinical relevance and previous literature and categorized into patient‐ and graft‐level factors. Three models were developed, adjusting for patient‐level variables, graft‐level variables, and both sets of variables combined, respectively. An additional sensitivity analysis further adjusted for surgeon identity to account for potential surgeon‐level clustering. Risk estimates are reported as adjusted odds ratios (aORs) with 95% CIs.

As a sensitivity analysis, propensity score matching was performed using all baseline patient‐ and graft‐level covariates. One‐to‐one nearest‐neighbor matching without replacement was applied using a caliper width of 0.15, and covariate balance was assessed using standardized mean differences. In addition, because LIMA patency was repeatedly assessed at 3 months, 1 year, and 3 years, an additional longitudinal sensitivity analysis using generalized estimating equations was performed to account for within‐patient correlation. An interaction between RAS group and follow‐up time was tested to assess whether the association between anastomotic location and graft occlusion varied over time, with adjustment for clinically relevant covariates.

Receiver operating characteristic analysis was used to determine the exploratory RAS threshold for predicting LIMA occlusion based on the maximum Youden index. Internal validation was performed using 1000 bootstrap resamples to assess threshold stability and the robustness of the observed association. Restricted cubic spline analyses with 3 knots were additionally performed to evaluate the continuous association between RAS and graft occlusion.

For the end point of MACEs, 3 Cox proportional hazards models were built to adjust for patient‐level, vessel‐level, and combined variables as mentioned above. Risk estimates of MACEs were presented as adjusted hazard ratio (aHR) and corresponding 95% CI.

Missing data were addressed using 5‐fold multiple imputation by chained equations (mice package, R software), and estimates were combined according to Rubin rules. 8

Statistical analyses were performed using R software, version 4.3.1 (R Foundation). P<0.05 was considered statistically significant.

RESULTS

Relation Between RAS and LIMA Occlusion

A total of 1597 patients with isolated proximal LAD lesions undergoing LIMA‐to‐LAD grafting were included in the present study. The mean age of the cohort was 60.4 years, and 80.9% were men. The overall occlusion rates of LIMA at 3 months, 1 year, and 3 years were 1.9%, 2.8%, and 4.3%, respectively. At 3 years, LIMAs that occluded had significantly lower median RAS values compared with those that remained patent (42% [38.2%–47.6%] versus 46.2% [39.7%–53.1%]; P=0.005).

To explore the relationship between RAS and LIMA occlusion, the RAS was categorized into 4 groups=<40%, 40% to 50%, 50% to 60%, and >60%. As RAS increased, corresponding to a more distal LAD anastomotic site, the LIMA occlusion rate showed a decreasing trend (Figure 3). When RAS was analyzed as a continuous variable using restricted cubic spline modeling, the risk of LIMA occlusion showed an overall decreasing trend with increasing RAS values (Figure S2).

Figure 3. The distribution of LIMA occlusion rates across RAS intervals.

Figure 3

LIMA indicates left internal mammary artery; and RAS, relative anastomotic site.

Receiver operating characteristic analysis identified 45.9% as the exploratory RAS cutoff for predicting LIMA occlusion, and thus a threshold of 46% was used for subsequent analyses (Figure S3). Internal validation using 1000 bootstrap resamples demonstrated good threshold stability, with a median bootstrap‐derived cutoff of 45.9% (interquartile range, 45.4%–45.9%).

Validation of the RAS Measurement

In the reproducibility analysis, the interobserver intraclass correlation coefficient was 0.86 (95% CI, 0.60–0.94), and the intraobserver intraclass correlation coefficient was 0.89 (95% CI, 0.81–0.94), indicating good measurement reproducibility.

In the anatomic validation cohort, patients classified as having distal anastomoses by the RAS method (RAS >46%) demonstrated significantly more distal anastomotic locations when measured directly on reconstructed CCTA images.

The actual anastomotic distance from the left coronary origin was greater in the distal group than in the proximal group (85.6±11.1 versus 72.4±12.9 mm; P<0.001). After normalization for total LAD length, the actual relative anastomotic position remained significantly more distal (54.4%±10.2% versus 47.6%±12.6%; P<0.001), whereas total LAD length was similar between groups (160.0±24.2 versus 158.0±32.9 mm). These findings support the anatomic validity of the RAS metric as a surrogate measure of the relative location of the LIMA‐LAD anastomosis along the LAD.

Baseline Characteristics Based on RAS Threshold

Patients were divided into proximal (RAS ≤46%) and distal (RAS >46%) anastomotic groups. Those with distal anastomosis had significantly higher RAS values (54.1%±6.1% versus 38.5%±5.8%; P<0.001). Table 1 summarizes the differences in patient‐ and vessel‐level characteristics. Compared with the proximal group, patients in the distal group had a slightly lower body mass index and higher SYNTAX (The SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery) scores. They were more frequently men, had a higher prevalence of prior myocardial infarction, and were more often operated on with on‐pump CABG surgery. Although the proportion of high‐volume surgeons was slightly higher in the proximal group, the saphenous vein graft occlusion rates were similar between the 2 groups (12.1% versus 10.1%; P=0.23), suggesting comparable surgical proficiency across both groups.

Table 1.

Comparison of Patient‐ and Vessel‐Level Characteristics According to RAS of LIMA‐LAD anastomosis

Characteristic Proximal group (RAS ≤46%) Distal group (RAS >46%) P value
(N=803) (N=794)
Patient characteristics
Male sex 629 (78.3) 663 (83.5) 0.01
Age, y 60.2±8.3 60.6±8.1 0.351
BMI, kg/m2 25.9±3.3 25.4±3.0 <0.001
LVEF <50% 53 (6.6) 52 (6.5) >0.99
NYHA class III/IV 322 (40.1) 359 (45.2) 0.044
Smoking 460 (57.3) 454 (57.2) >0.99
Diabetes 295 (36.7) 311 (39.2) 0.34
Hypertension 481 (59.9) 478 (60.2) 0.94
Hypercholesterolemia 613 (76.3) 620 (78.1) 0.44
PVD 20 (2.5) 16 (2.0) 0.64
Prior MI 135 (16.8) 184 (23.2) 0.002
3‐Vessel disease 699 (87.0) 704 (88.7) 0.36
Left main disease 264 (32.9) 273 (34.4) 0.56
SYNTAX score 29.8±9.28 31.4±9.57 <0.001
On‐pump surgery 284 (35.4) 371 (46.7) <0.001
High‐volume surgeon (>100 CABGs per year) 601 (74.8) 543 (68.4) 0.005
Patient‐level SVG occlusion 97 (12.1) 80 (10.1) 0.23
Vessel characteristics
RAS, % 39±6 54±6 <0.001
Anastomosis height, mm 53.2±16.1 75.2±21.1 <0.001
Heart height, mm 138±35.6 139±34.8 0.71
QFR >0.80 69 (8.6) 51 (6.4) 0.123
Degree of stenosis, % 87.0±11.3 87.7±10.7 0.17
LAD occlusion 219 (27.3) 223 (28.1) 0.76
Prior PCI of LAD 56 (7.0) 56 (7.1) >0.99
Run‐offs* 1.62±0.26 1.56±0.24 <0.001
Pulsatility index 2.34±0.80 2.21±0.67 0.003
Mean graft flow, mL/min 26 (19–36) 29 (20–40) 0.002

Data are given as mean±SD, number (percentage), or mean (interquartile range). BMI indicates body mass index; CABG, coronary artery bypass grafting; LAD, left anterior descending artery; LIMA, left internal mammary artery; LVEF, left ventricular ejection fraction; MI, myocardial infarction; SYNTAX, the SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery; NYHA, New York Heart Association; PCI, percutaneous coronary intervention; PVD, peripheral vascular disease; QFR, quantitative flow ratio; RAS, relative anastomotic site; and SVG, saphenous vein graft.

*

Run‐off referred to the intraoperative assessment of the LAD target vessel caliber at the planned anastomotic site.

At the vessel level, patients with distal LIMA anastomosis exhibited significantly smaller run‐offs, lower pulsatility index, and higher mean flow of LIMA graft (Table 1).

Univariable and Multivariable Analysis

At both 1 and 3 years postoperatively, distal LIMA anastomosis was associated with significantly lower occlusion rates compared with proximal anastomosis (1.4% versus 4.1%, P=0.002; 2.3% versus 6.3%, P<0.001, respectively) (Figure 4).

Figure 4. Comparison of LIMA occlusion rates between distal and proximal anastomosis.

Figure 4

LIMA indicates left internal mammary artery; and RAS, relative anastomotic site. *P<0.05, **P<0.001.

Using 3‐year LIMA occlusion as the end point, 3 multivariable logistic regression models were constructed to adjust for patient‐level variables, vessel‐level variables, and combined variables. Distal anastomotic location (RAS >46%) was a protective factor against 3‐year LIMA occlusion in all 3 models (Table 2). Additional adjustment for surgeon identity yielded results consistent with the primary analysis, with distal LIMA‐LAD anastomosis remaining independently associated with a lower risk of 3‐year graft occlusion.

Table 2.

Multivariable Analysis of the Impact of RAS on 3‐Year LIMA Occlusion Rates

Model aOR for RAS >46% 95% CI P value
Model 1* 0.36 0.20–0.65 <0.001
Model 2† 0.37 0.20–0.66 <0.001
Model 3‡ 0.37 0.20–0.69 0.002

aOR indicates adjusted odds ratio; LIMA, left internal mammary artery; and RAS, relative anastomotic site.

*

Model 1 adjusted for patient‐level variables.

†

Model 2 adjusted for vessel‐level variables.

‡

Model 3 adjusted for all the variables (detailed in the Methods section).

Sensitivity Analysis Using Propensity Score Matching

To further assess the robustness of the findings and minimize residual confounding, a propensity score matching analysis was performed. Because of missing values in variables used for propensity score estimation, 121 patients were excluded from the matching procedure. Before matching, 745 patients in the proximal anastomosis group and 731 patients in the distal anastomosis group were available for analysis. After 1:1 nearest‐neighbor matching, 618 matched pairs were generated.

Several baseline imbalances were observed before matching, including sex, prior myocardial infarction, SYNTAX score, on‐pump surgery LAD run‐off, and LIMA pulsatility index (PI). After matching, all covariates achieved satisfactory balance (Table S1).

In the matched cohort, distal LIMA‐LAD anastomosis remained significantly associated with a lower risk of 3‐year LIMA occlusion (odds ratio, 0.37 [95% CI, 0.19–0.69]; P=0.003), consistent with the primary analysis. These findings support the robustness of the observed association between a more distal anastomotic location and improved LIMA graft patency.

Sensitive Analysis Using Generalized Estimating Equations

In the longitudinal generalized estimating equation analysis incorporating LIMA patency assessments at 3 months, 1 year, and 3 years, there was no significant interaction between RAS group and follow‐up time (P for interaction=0.29), indicating no evidence that the association between anastomotic location and graft occlusion differed across follow‐up periods. In the model without the interaction term, distal LIMA‐LAD anastomosis remained associated with a lower overall risk of graft occlusion (aOR, 0.33 [95% CI, 0.17–0.63]; P<0.001), consistent with the primary 3‐year analysis.

Patient Outcomes

At 3 years postoperatively, the overall MACE rate was 3.9%. MACE rates between patients with distal and proximal location of LIMA anastomosis were similar (3.4% versus 4.4%) (Figure S4). Three multivariable Cox regression models were constructed to adjust for patient‐level variables, vessel‐level variables, and combined variables. RAS was not significantly associated with MACEs in all models (Table S2). However, patients with distal anastomosis had significantly lower rates of repeat revascularization compared with the proximal group (1.0% versus 2.6%; aHR, 0.30 [95% CI, 0.12–0.87]; P=0.014) (Table 3).

Table 3.

Comparison of Clinical Outcomes Between Patients With Proximal and Distal Location of LIMA‐LAD Anastomosis

Outcome RAS ≤46% RAS >46% aHR (95% CI) for RAS >46% P value
(N=803), No. (%) (N=794), No. (%)
MACEs 35 (4.4) 27 (3.4) 0.68 (0.40–1.15) 0.16
All‐cause death 10 (1.3) 15 (1.9) 1.42 (0.58–3.48) 0.44
Nonfatal MI 13 (1.6) 5 (0.6) 0.31 (0.09–1.02) 0.054
Repeat revascularization 21 (2.6) 8 (1.0) 0.30 (0.12–0.87) 0.014*

All patient‐ and vessel‐level variables were adjusted. aHR indicates adjusted hazard ratio; LAD, left anterior descending artery; LIMA, left internal mammary artery; MACE, major adverse cardiac event; MI, myocardial infarction; and RAS, relative anastomotic site.

*

P < 0.05.

LIMA occlusion was strongly associated with adverse outcomes. Patients with LIMA occlusion experienced a significantly higher MACE rate at 3 years compared with those that remained patent (13.2% versus 3.5%; aHR, 3.74 [95% CI, 1.80–7.79]; P=0.001) (Figure S5).

DISCUSSION

In this secondary analysis of the patient cohort from the PATENCY trial, a distal location of LIMA‐to‐LAD anastomosis (RAS >46%) was associated with significantly lower occlusion rates of LIMA at 3 years postoperatively. Although the overall rate of MACEs at 3 years was comparable between proximal and distal anastomotic groups, patients with distal anastomoses might be associated with a lower incidence of repeat revascularization.

A major strength of the present study is the development of a quantitative imaging‐based approach for assessing the location of the LIMA‐LAD anastomosis. The optimal anastomotic site for patients with isolated proximal LAD disease remains uncertain, as both proximal and more distal targets have potential advantages. Leveraging the systematic CCTA follow‐up available in the PATENCY data set, we sought to evaluate this question using a standardized anatomic metric.

The RAS metric was derived using 2 reproducible anatomic landmarks: the pulmonary sinus and the cardiac apex. The pulmonary sinus was selected because it is readily identifiable both on CCTA and during CABG surgery, and lies at a similar anatomic level to the left main coronary artery origin, facilitating translation of imaging findings into a clinically meaningful surgical context. The cardiac apex was chosen as the distal reference point because it represents a consistent anatomic landmark and allows normalization for differences in heart size. By relating the anastomotic position to the distance between these 2 landmarks, the RAS metric provides a standardized assessment of relative anastomotic location that is less dependent on individual cardiac dimensions.

Potential Explanations

Several physiological and anatomic factors may potentially contribute to the observed association between a more distal LIMA–LAD anastomotic location and lower graft occlusion risk; however, these mechanisms were not directly evaluated in the present study.

Competitive flow may be 1 possible explanation, as it is a recognized determinant of arterial graft patency. 4 , 9 , 10 , 11 , 12 In the present study, distal anastomoses were associated with smaller target‐vessel caliber, which may reflect different competitive flow conditions, and prior experimental studies have reported higher graft flow with more distal anastomoses under competitive flow. 13 On the other hand, differences in graft geometry or anastomotic configuration may also influence local flow patterns and endothelial stress, as proximal LIMA–LAD anastomoses may involve a more perpendicular anastomotic angle, whereas more distal anastomoses may create a more favorable anastomotic angle and smoother blood flow. Nevertheless, anastomotic angle and detailed graft geometry were not systematically assessed in the present study, and therefore these considerations should be regarded as hypothesis generating rather than mechanistic conclusions.

Clinical Implications

Although the study found a lower MACE rate in patients with distal anastomoses, the difference was not statistically significant. This outcome may be attributable to the modest absolute difference in LIMA occlusion rates (4.0%) and the ability of residual native coronary flow to maintain myocardial perfusion even after graft occlusion, at least in the short term. 14 , 15 However, LIMA occlusion was significantly associated with adverse clinical outcomes, and patients with distal anastomoses experienced fewer repeat revascularization events. These findings align with prior studies showing that graft failure is linked to increased risks of myocardial infarction and repeat interventions. 3 , 16 Over time, as native coronary artery disease progresses, the differences in outcomes between groups may become more pronounced, underscoring the need for longer follow‐up to fully evaluate the impact of anastomotic location.

Limitations

Several limitations should be acknowledged when interpreting our results. First, because of the retrospective nature, unmeasured confounders associated with LIMA occlusion may exist and cannot be adjusted. Second, selection bias cannot be ruled out, as 388 eligible patients were excluded because of missing preoperative angiography or postoperative CCTA data. Third, because the primary end point (LIMA occlusion) and anastomotic site were determined using CCTA, the researchers assessing the images were not blinded to the study end points. Fourth, because the number of repeat revascularization events was limited, particularly in the distal anastomosis group, the statistical power for this secondary end point was limited and the findings should be interpreted cautiously. Fifth, the present study included only an Asian CABG population, and its generalizability to populations with different ethnic and anatomic characteristics remains uncertain and requires validation in independent external cohorts.

CONCLUSIONS

This study suggests that a more distal LIMA‐to‐LAD anastomotic location may be associated with improved 3‐year graft patency. These findings warrant further prospective trials to confirm the long‐term effects of anastomotic location on both graft patency and patient outcomes.

Sources of Funding

This study was supported by National High Level Hospital Clinical Research Funding (2023‐GSP‐RC‐03 and 2024‐GSP‐TJ‐01), Beijing Natural Science Foundation—Fengtai Innovation Joint Fund (L251022), General Program of the Natural Science Foundation of China (grants 82570411, 82370401, and 82470353), and the Young Scientists Fund (Category C) of the Natural Science Foundation of China (grant 82600694).

Disclosures

The authors report no relevant financial relationships, commercial affiliations, or conflicts of interest related to this work.

Supporting information

Tables S1–S2

Figures S1–S5

JAH3-15-e051333-s002.pdf (585.8KB, pdf)

STROBE checklist

JAH3-15-e051333-s001.pdf (423.3KB, pdf)

This manuscript was sent to John S. Ikonomidis, MD, PhD, Guest Editor, for review by expert referees, editorial decision, and final disposition.

For Sources of Funding and Disclosures, see page 8 and 9.

Contributor Information

Yan Zhang, Email: waikezhangyan@126.com.

Zhan Hu, Email: huzhan@fuwai.com.

References

  • 1. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW, Golding LA, Gill CC, Taylor PC, Sheldon WC. Influence of the internal‐mammary‐artery graft on 10‐year survival and other cardiac events. N Engl J Med. 1986;314:1–6. doi: 10.1056/NEJM198601023140101 [DOI] [PubMed] [Google Scholar]
  • 2. Zeff RH, Kongtahworn C, Iannone LA, Gordon DF, Brown TM, Phillips SJ, Skinner JR, Spector M. Internal mammary artery versus saphenous vein graft to the left anterior descending coronary artery: prospective randomized study with 10‐year follow‐up. Ann Thorac Surg. 1988;45:533–536. doi: 10.1016/s0003-4975(10)64526-2 [DOI] [PubMed] [Google Scholar]
  • 3. Gaudino M, Sandner S, An KR, Dimagli A, Di Franco A, Audisio K, Harik L, Perezgrovas‐Olaria R, Soletti G, Fremes SE, et al. Graft failure after coronary artery bypass grafting and its association with patient characteristics and clinical events: a pooled individual patient data analysis of clinical trials with imaging follow‐up. Circulation. 2023;148:1305–1315. doi: 10.1161/CIRCULATIONAHA.123.064090 [DOI] [PubMed] [Google Scholar]
  • 4. Wang C, Hu Z, Hou Z, Wang Y, Song L, Xu B, Guan C, Ning Y, Feng W, Zhang Y. Impact of preoperative quantitative flow ratio of the left anterior descending artery on internal mammary artery graft patency and midterm patient outcomes after coronary artery bypass grafting. J Am Heart Assoc. 2023;12:e029134. doi: 10.1161/JAHA.122.029134 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Tian M, Wang X, Sun H, Feng W, Song Y, Lu F, Wang L, Wang Y, Xu B, Wang H, et al. No‐touch versus conventional vein harvesting techniques at 12 months after coronary artery bypass grafting surgery: multicenter randomized, controlled trial. Circulation. 2021;144:1120–1129. doi: 10.1161/CIRCULATIONAHA.121.055525 [DOI] [PubMed] [Google Scholar]
  • 6. Wang X, Tian M, Zheng Z, Gao H, Wang Y, Wang L, Hu S. Rationale and design of a multicenter randomized trial to compare the graft patency between no‐touch vein harvesting technique and conventional approach in coronary artery bypass graft surgery. Am Heart J. 2019;210:75–80. doi: 10.1016/j.ahj.2018.11.011 [DOI] [PubMed] [Google Scholar]
  • 7. FitzGibbon GM, Burton JR, Leach AJ. Coronary bypass graft fate: angiographic grading of 1400 consecutive grafts early after operation and of 1132 after one year. Circulation. 1978;57:1070–1074. doi: 10.1161/01.cir.57.6.1070 [DOI] [PubMed] [Google Scholar]
  • 8. Rubin DB. Multiple Imputation for Nonresponse in Surveys. Wiley; 1987. doi: 10.1002/9780470316696 [DOI] [Google Scholar]
  • 9. Berger A, MacCarthy PA, Siebert U, Carlier S, Wijns W, Heyndrickx G, Bartunek J, Vanermen H, De Bruyne B. Long‐term patency of internal mammary artery bypass grafts: relationship with preoperative severity of the native coronary artery stenosis. Circulation. 2004;110:II36–II40. doi: 10.1161/01.CIR.0000141256.05740.69 [DOI] [PubMed] [Google Scholar]
  • 10. Sabik JF, Lytle BW, Blackstone EH, Houghtaling PL, Cosgrove DM. Comparison of saphenous vein and internal thoracic artery graft patency by coronary system. Ann Thorac Surg. 2005;79:544–551. doi: 10.1016/j.athoracsur.2004.07.047 [DOI] [PubMed] [Google Scholar]
  • 11. Harskamp RE, Alexander JH, Ferguson TB, Hager R, Mack MJ, Englum B, Wojdyla D, Schulte PJ, Kouchoukos NT, de Winter RJ, et al. Frequency and predictors of internal mammary artery graft failure and subsequent clinical outcomes: insights from the project of ex‐vivo vein graft engineering via transfection (PREVENT) IV trial. Circulation. 2016;133:131–138. doi: 10.1161/CIRCULATIONAHA.115.015549 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Glineur D, Grau JB, Etienne P‐Y, Benedetto U, Fortier JH, Papadatos S, Laruelle C, Pieters D, El Khoury E, Blouard P, et al. Impact of preoperative fractional flow reserve on arterial bypass graft anastomotic function: the IMPAG trial. Eur Heart J. 2019;40:2421–2428. doi: 10.1093/eurheartj/ehz329 [DOI] [PubMed] [Google Scholar]
  • 13. Jackson ZS, Ishibashi H, Gotlieb AI, Langille BL. Effects of anastomotic angle on vascular tissue responses at end‐to‐side arterial grafts. J Vasc Surg. 2001;34:300–307. doi: 10.1067/mva.2001.115815 [DOI] [PubMed] [Google Scholar]
  • 14. Paterson HS, Bannon PG, Taggart DP. Competitive flow in coronary bypass surgery: the roles of fractional flow reserve and arterial graft configuration. J Thorac Cardiovasc Surg. 2017;154:1570–1575. doi: 10.1016/j.jtcvs.2017.05.087 [DOI] [PubMed] [Google Scholar]
  • 15. Gaudino M, Di Franco A, Bhatt DL, Alexander JH, Abbate A, Azzalini L, Sandner S, Sharma G, Rao SV, Crea F, et al. The association between coronary graft patency and clinical status in patients with coronary artery disease. Eur Heart J. 2021;42:1433–1441. doi: 10.1093/eurheartj/ehab096 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Malinowski M, Mrozek R, Twardowski R, Biernat J, Deja MA, Widenka K, Dalecka A‐M, Kobielusz‐Gembala I, Janusiewicz P, Wos S, et al. Left internal mammary artery improves 5‐year survival in patients under 40 subjected to surgical revascularization. Heart Surg Forum. 2006;9:E493–E497. doi: 10.1532/HSF98.20041182 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Tables S1–S2

Figures S1–S5

JAH3-15-e051333-s002.pdf (585.8KB, pdf)

STROBE checklist

JAH3-15-e051333-s001.pdf (423.3KB, pdf)

Articles from Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease are provided here courtesy of Wiley

RESOURCES