Introduction
Microsurgical technique has made it possible to reconstruct various vessels and perform complete myocardial revascularization even in high-risk patients with multiple occlusions, diffuse lesions and previous stenting [1]. Although anastomoses imposed using microsurgical technique usually provide clinical success and favorable early postoperative outcomes, some surgeons still question the need for bypass of small coronary arteries (<1.5 mm). However, comparative analysis of postoperative results in these patients has not yet been carried out [2, 3].
To study the effectiveness of complete microsurgical myocardial revascularization, we analyzed in-hospital results of coronary artery bypass grafting (CABG) in patients with coronary artery diameter < 1.5 mm compared to ones with larger vessels in homogeneous groups. Propensity score matching was applied for significant confounders.
Material and methods
A single-center prospective study with retrospective propensity score matching was performed. This trial enrolled all patients with angina CCS class II-IV and multiple-vessel coronary artery disease who underwent elective CABG between January 2018 and December 2020. We excluded patients with severe chronic heart failure and severe pulmonary hypertension, low left ventricular ejection fraction (<35%), severe comorbidities, myocardial infarction within previous 1.5 months, combined or previous cardiac and vascular surgeries. We also excluded off-pump CABG.
Thus, we recruited 460 patients and divided them into 2 groups. The main group included patients with multiple lesions of ≥ 2 coronary arteries with a diameter of less than 1.5 mm within the anastomosis site (group 1, n=105). The control group consisted of patients with a target artery diameter of more than 1.5 mm within the anastomosis site (group 2, n=355).
Selected patients were analyzed regarding 18 confounders. Patients in the main group were older (65.0±9.1 and 62.5±8.1 years, p=0.007). Moreover, patients with angina NYHA class 3 (82 (78.1%) vs. 240 (67.6%), p=0.040), chronic obstructive pulmonary disease (12 (11.4%) vs. 14 (3.9%), p=0.007), multifocal atherosclerosis (47 (44.8%) vs. 103 (29.0%), p=0.003), previous stroke (10 (9.5%) vs. 11 (3, 1%), p=0.013) and chronic kidney disease stage ≤3A (7 (6.7%) vs. 7 (2.0%), p=0.022) prevailed in the same group. High Syntax score (>32) (85 (81.0%) vs. 206 (58.0%), p<0.001), STS score (0.8±0.4 vs 0 .7±0.3, p=0.006) and lower glomerular filtration rate (84.9±19.6 vs. 94±20.3 ml/min, p<0.001) were also common in this group. Other parameters were similar.
Considering between-group heterogeneity, we performed propensity score matching. The minimum estimated size of each group was 100 patients according to the Lehr’s formula. Both groups were analyzed according to 18 parameters (confounders). The results are summarized in Table 1.
Table 1. Clinical characteristics of patients after propensity score matching
Variable |
Group 1 (n=100) |
Group 2 (n=100) |
p-value |
Males, n (%) |
76 (76.0) |
82 (82.0) |
0.386 |
Age, years |
65±9.2 |
65±8.0 |
1.000 |
Obesity (BMI>30 kg/m2), n (%) |
28 (28.0) |
33 (33.0) |
0.539 |
Smoking, n (%) |
36 (36.0) |
40 (40.0) |
0.662 |
Angina CCS class II, n (%) |
6 (6.0) |
8 (8.0) |
0.783 |
Angina CCS class III, n (%) |
78 (78.0) |
69 (69.0) |
0.200 |
Angina CCS class IV, n (%) |
13 (13.0) |
15 (15.0) |
0.839 |
Unstable angina, n (%) |
1 (1.0) |
1 (1.0) |
1.000 |
Silent ischemia, n (%) |
2 (2.0) |
7 (7.0) |
0.170 |
Arterial hypertension, n (%) |
84 (84.0) |
91 (91.0) |
0.199 |
Diabetes mellitus, n (%) |
33 (33.0) |
33 (33.0) |
1.000 |
COPD stage I, n (%) |
12 (12.0) |
6 (6.0) |
0.216 |
Multifocal atherosclerosis, n (%) |
42 (42.0) |
44 (44.0) |
0.887 |
Previous stroke, n (%) |
10 (10.0) |
11 (11.0) |
1.000 |
Post-infarction cardiac sclerosis, n (%) |
51 (51.0) |
56 (56.0) |
0.571 |
Previous PCI, n (%) |
22 (22.0) |
24 (24.0) |
0.867 |
Left ventricular ejection fraction, % |
60±5.2 |
60±5.9 |
1.000 |
Heart failure NYHA class I-II, n (%) |
18 (18.0) |
22 (22.0) |
0.596 |
GFR, ml/min |
83.6±19.4 |
86.8±17.5 |
0.222 |
Chronic kidney disease stage ≤ 3A, n (%) |
5 (5.0) |
5 (5.0) |
1.000 |
Syntax score >32, n (%) |
85 (85.0) |
76 (76.0) |
0.153 |
STS score |
0.8±0.4 |
0.7±0.5 |
0.120 |
Note. BMI — body mass index, COPD — chronic obstructive pulmonary disease, PCI — percutaneous coronary intervention, GFR — glomerular filtration rate.
Surgical approach
Two teams of experienced surgeons performed all surgeries under endotracheal anesthesia through midline sternotomy. Moderate hypothermia (34°C), cardiopulmonary bypass and crystalloid cardioplegia (Custodiol) were applied. An operating microscope and 8-9/0 sutures with 6.5-mm needles were used for distal anastomoses. All patients underwent complete anatomical myocardial revascularization, i.e. bypass and reconstruction of at least 1 coronary artery in each vascular system (circumflex artery, left anterior descending artery, right coronary artery) regardless severity of arterial damage (anastomoses with arteries < 1.5 mm (up to 0.5 mm)) with simultaneous endarterectomy for vascular occlusion or calcification.
The left internal mammary artery (LIMA) was routinely used for anastomosis with LAD. The right internal mammary artery (RIMA) was used according to indications in patients younger 60 years old without obesity and diabetes mellitus. Great and small saphenous vein grafts were also harvested. Radial arteries were used if venous grafts were inappropriate for CABG.
Dual antiplatelet therapy with aspirin and clopidogrel was prescribed after reconstruction of coronary arteries < 1.5 mm within the anastomosis site or in case of extended anastomoses. Other patients received standard monotherapy with acetylsalicylic acid. To prevent postoperative thromboembolic complications, we additionally prescribed enoxaparin sodium for 2-4 days. Coronary endarterectomy required subsequent 6-month anticoagulation with warfarin in addition to aspirin therapy under control of international normalized ratio (target values 2.0-3.0).
Results
Total and mean number of distal anastomoses was higher in the main group. However, there was no significant between-group difference (Table 2). The number of IMA grafts was approximately the same. At the same time, radial arteries were used in 18 patients with coronary arteries ≤ 1.5 mm due to the poor quality of venous grafts. The main group was characterized by significantly more common sequential and Y-grafts (57/418 vs. 26/371), prolonged angioplasty (37/418 vs. 6/371) and endarterectomy (19/418 vs. 3/371).
Table 2. Surgical characteristics
Variable |
Group 1 (n=100) |
Group 2 (n=100) |
p-value |
LIMA — LAD, n (%) |
98 (98.0) |
99 (99.0) |
1.000 |
RIMA grafting, n (%) |
17 (17.0) |
14 (14.0) |
0.697 |
Radial artery grafting, n (%) |
18 (18.0) |
0 |
<0.001 |
Mean number of distal anastomoses, n |
4.0±1.1 |
4.0±0.8 |
1.000 |
Total number of distal anastomoses, n |
418 |
371 |
|
Standard anastomoses with coronary arteries > 1,5 mm, n (%) |
61/418 (14.6) |
366/371 (98.7) |
<0.001 |
Distal shunting of coronary artery ≤1,5 mm, n (%) |
357/418 (85.4) |
5/371 (1.3) |
<0.001 |
Distal shunting of coronary artery ≤ 1 mm from the total number of bypassed arteries ≤1,5 mm, n (%) |
131/357 (36.7) |
5/5 (100) |
0.007 |
Sequential anastomoses and Y-grafting, n (%) |
57/418 (13.6) |
26/371 (7.0) |
0.003 |
Extended (> 20 mm) graft-related angioplasty, n (%) |
37/418 (8.9) |
6/371 (1.6) |
<0.001 |
Endarterectomy, n (%) |
19/418 (4.5) |
3/371 (0.8) |
0.002 |
Note. LIMA — left internal mammary artery, LAD — left anterior descending artery, RIMA — right internal mammary artery.
Analyzing intraoperative parameters in the main group, we found significantly worse outcomes mainly for anastomoses with coronary arteries ≤ 1 mm. Therefore, we distinguished a subgroup of patients with lesions of ≥ 2 coronary arteries with a diameter ≤ 1 mm to obtain objective data.
Higher incidence of complex techniques (sequential anastomoses, endarterectomy) in the primary group was followed by prolonged cardiopulmonary bypass and aortic cross-clamping (Table 3). Intraoperative blood loss was also significantly higher in the main group. Perhaps, this is due to previous aggressive antithrombotic therapy and subsequent need for blood transfusion (without statistical significance).
Table 3. Intraoperative characteristics
Variable |
Group 1 (n=100) |
Group 2 (n=100) |
p-value |
|
CA ≤1,5 mm (n=61) |
CA ≤1 mm (n=39) |
|||
Aortic cross-clamping time, min |
65±21 |
65±45 |
57±16 |
p1—2=1.000 p1—3=0.007 p2—3=0.123 |
Cardiopulmonary bypass time, min |
95±28 |
95±78 |
86±24 |
p1—2=1.000 p1—3=0.032 p2—3=0.301 |
Blood transfusion, n (%) |
2 (3.3) |
12 (30.8) |
8 (8.0) |
p1—2<0.001 p1—3=0.321 p2—3=0.002 |
Replenished blood loss, ml |
520±91 |
526±92 |
390±25 |
p1—2=0.750 p1—3<0.001 p2—3<0.001 |
Sternotomy for bleeding, n (%) |
0 |
0 |
1 (1.0) |
1.000 |
Need for inotropic/vasopressor support, n (%) |
0 |
2 (5.1) |
3 (3.0) |
p1—2=0.150 p1—3=0.290 p2—3=0.628 |
Studying the outcomes in patients with coronary arteries ≤ 1 mm, we found significantly longer time of cardiopulmonary bypass and aortic cross-clamping, as well as blood loss in comparison with patients with coronary arteries ≤ 1.5 mm and the control group.
Mean postoperative hospital-day was higher in patients with small-diameter coronary arteries. One patient with coronary artery diameter ≤ 1 mm had perioperative myocardial infarction with subsequent fatal outcome. Incidence of other adverse events, such as stroke, acute renal failure, cardiac arrhythmias, was similar in both groups (Table 4).
Table 4. In-hospital results
Variable |
Group 1 (n=100) |
Group 2 (n=100) |
p-value |
|
CA ≤1,5 mm (n=61) |
CA ≤1 mm (n=39) |
|||
ICU-stay, days |
2±0.6 |
2±0.6 |
2±0.4 |
1.000 |
Postoperative hospital-day, days |
11±1.8 |
11±2 |
10±2 |
p1—2=1.000 p1—3=0.002 p2—3=0.009 |
In-hospital mortality, n (%) |
0 |
1 (2.6) |
0 |
p1—2=0.390 p1—3=1.000 p2—3=0.281 |
Perioperative myocardial infarction, n (%) |
0 |
1 (2.6) |
0 |
p1—2=0.390 p1—3=1.000 p2—3=0.281 |
Stroke, n (%) |
0 |
0 |
2 (2.0) |
p1—2=1.000 p1—3=0.526 p2—3=1.000 |
Paroxysmal atrial fibrillation de novo, n (%) |
7 (11.5) |
8 (20.5) |
16 (16.0) |
p1—2=0.257 p1—3=0.493 p2—3=0.618 |
Renal failure (plasmapheresis), n (%) |
1 (1.6) |
0 |
1 (1.0) |
1.000 |
Note. CA — coronary artery, ICU — intensive care unit.
All patients in the main group were scheduled for CT angiography within 1 month after surgery. With the exception of 1 patient with a fatal outcome, 99 CT angiography procedures were performed. These data are summarized in Table 5. Graft dysfunction was defined as stenosis > 50% or occlusion in any site. Moreover, diffuse IMA narrowing <1 mm (string sign) was also classified as graft dysfunction [4].
Table 5. Postoperative angiography data in the main group
CA diameter |
Autograft |
Anastomosis |
Graft function |
|||
RCA |
Cx/a.intermedia |
LAD |
Patent |
Failed |
||
≤1,5 mm (n=226) |
LIMA, n (%) |
— |
2 (0.9) |
60 (26.5) |
62 (27.4) |
— |
RIMA, n (%) |
5 (2.2) |
4 (1.8) |
3 (1.3) |
12 (5.3) |
— |
|
Radial artery, n (%) |
2 (0.9) |
6 (2.7) |
4 (1.8) |
12 (5.3) |
— |
|
Saphenous vein, n (%) |
49 (21.7) |
62 (27.4) |
29 (12.8) |
138 (61.1) |
2 (0.9) |
|
≤1 mm (n=127) |
LIMA, n (%) |
— |
1 (0.8) |
38 (29.1) |
38 (29.9) |
1 (0.8) |
RIMA, n (%) |
1 (0.8) |
— |
1 (0.8) |
1 (0.8) |
1 (0.8) |
|
Radial artery, n (%) |
3 (2.4) |
— |
1 (0.8) |
4 (3.1) |
— |
|
Saphenous vein, n (%) |
34 (26.8) |
29 (22.8) |
19 (15.0) |
76 (59.8) |
6 (4.7) |
|
>1,5 mm (n=61) |
LIMA, n (%) |
— |
— |
9 (14.8) |
9 (14.8) |
— |
RIMA, n (%) |
1 (1.6) |
1 (1.6) |
2 (3.3) |
4 (6.6) |
— |
|
Radial artery, n (%) |
— |
3 (4.9) |
— |
3 (4.9) |
— |
|
Saphenous vein, n (%) |
10 (16.4) |
26 (42.6) |
9 (14.8) |
45 (73.8) |
— |
Note. LIMA — left internal mammary artery, RIMA — right internal mammary artery, RCA — right coronary artery, Cx — circumflex artery, LAD — left anterior descending artery.
There were 2 occlusions of venous grafts to diagonal branch with a diameter of ≤ 1.5 mm and 8 occlusions of grafts to coronary arteries ≤ 1 mm (p=0.005). In 1 case, we observed contrast enhancement of the right internal mammary artery up to the distal third with subsequent competitive blood flow from the right coronary artery. In 1 case, there was occlusion of the second diagonal branch at the site of anastomosis and distal to this segment after sequential bypass grafting (IMA + LAD + diagonal branch). LAD was patent. Occlusion of saphenous vein grafts occurred in 6 cases including 2 patients with occlusion of one branch of the Y-graft.
Discussion
Constantly growing number of percutaneous coronary interventions and new opportunities of long-term therapy in patients with coronary artery disease have gradually changed the profile of patients referred for CABG [5]. Mean age of patients has increased in recent years. Multiple previous stenting procedures and long-standing diabetes mellitus became significantly more common. As a result, most candidates for CABG usually have multiple and extended coronary artery occlusions, diffuse distal lesions and “poor” peripheral coronary vessels [6]. These ones have already been denied surgery in regional hospitals as a rule.
Microsurgical technique used since the early 80s makes it possible to perform complete myocardial revascularization in patients with coronary arteries ≤ 1.5 mm [7, 8]. However, despite the obvious correlation of life expectancy and angina recurrence on the quality of revascularization [9, 10], most surgeons still doubt the advisability of surgery in patients with small arteries. They argue this approach by worse early and long-term outcomes [2, 3].
One of the causes of worse prognosis in these patients is high risk of graft stenosis and occlusion in the early and long-term postoperative period compared with those with “large” coronary arteries [11, 12]. Major predictors of graft dysfunction are imperfect technique of coronary anastomosis and worse blood flow through the anastomoses with small coronary arteries. You can compensate technical errors using advanced optical magnification and thin sutures [8, 13].
There was one postoperative myocardial infarction with subsequent fatal outcome among patients with coronary arteries ≤ 1 mm. The same group was characterized by higher incidence of graft occlusion in postoperative period. Nevertheless, survival and postoperative morbidity were similar in patients with coronary artery diameter less than and more than 1.5 mm in our study. Choosing surgical strategy, surgeon should consider potentially necessary complex techniques required for bypass of coronary arteries ≤ 1.5 mm (sequential and Y-grafting, extended anastomoses with angioplasty, single and multiple endarterectomies). These features increase the volume of surgery [14-17]. In our study, postoperative angiography in the main group revealed a more favorable function of arterial grafts compared to vein grafts that is also consistent with literature data [18-20].
In our study, we used dual antiplatelet therapy with aspirin and clopidogrel after reconstruction of coronary arteries < 1.5 mm or extended anastomoses. In case of endarterectomy, 6-month anticoagulation with warfarin was added to aspirin monotherapy. There are various modes of antithrombotic therapy in the foreign literature (from dual antiplatelet therapy with aspirin and clopidogrel with lifelong warfarin monotherapy after endarterectomy [20, 21] to the enhanced antithrombotic therapy after extended anastomoses with angioplasty [22]). According to the European Society of Cardiology guidelines (2019), the second antithrombotic drug in addition to aspirin is advisable in patients with high risk of ischemic events and low-to-moderate risk of bleeding [23, 24].
Thus, we first observed that complete myocardial revascularization in comparable groups of patients with coronary artery ≤ 1.5 mm provides favorable in-hospital outcomes comparable to the results of CABG in patients with larger arteries.
Conclusion
Complete myocardial revascularization is advisable in patients with coronary arteries ≤ 1.5 mm despite longer surgery and its technical complexity. At the same time, coronary artery bypass surgery in patients with coronary arteries ≤ 1 mm showed worse in-hospital outcomes and higher incidence of graft occlusion. Further study of this problem in long-term postoperative period is required for final conclusions.
The authors declare no conflicts of interest.