High mortality after coronary bypass surgery in patients with high-grade left main coronary artery stenosis

Abstract

Objective. To determine mortality after coronary artery bypass grafting (CABG) in relation to degree of left main coronary artery (LMCA) obstruction. Design. All patients without LMCA stenosis (n=3370), with low-grade stenosis (n = 261), high-grade stenosis (n = 224) or total occlusion of the LMCA (n = 15) were followed for ten years after CABG performed during 1970–1989. Results. Early mortality was 1.9% and 2.3%, respectively, if there was no or a low-grade LMCA stenosis vs. 6.3% if the stenosis was high-grade. Ten-year survival was 76% if no LMCA obstruction, 74% if low-grade stenosis and 64% if the stenosis was high-grade. Risk of early death (odds ratio 2.6, 95% CI 1.4–4.8) and mortality at ten years (relative risk 1.5, 95% CI 1.1–2.0) was higher in patients with high-grade stenosis than in those without LMCA stenosis. There was no increased long-term mortality in patients with low-grade stenosis or among the few patients with occlusion of the LMCA. Conclusions. High-grade LMCA stenosis was associated with a three-fold increased risk of early and fifty percent higher risk of late death than in patients without LMCA stenosis.

• Left main coronary artery
• occlusion
• survival
• coronary artery bypass
• collaterals

A prerequisite for survival in patients with a high-grade stenosis or occlusion left main coronary artery (LMCA) is gradual progress of the LMCA obstruction allowing enough time for the development of sufficient collateral circulation to the left ventricular myocardium. The development of coronary collaterals is a function of both the severity of the arterial stenosis and the level of antegrade flow in the epicardial vessels 1, 2. Chronic total occlusion of LMCA is an exclusive rare finding and the reported incidence in patients undergoing coronary artery bypass grafting (CABG) ranges from 0.04% up to 0.8% 3–7. Our hypothesis was that regression of collaterals occurring after successful coronary bypass surgery might put particularly patients with occlusion or a high-grade stenosis of the LMCA at risk in case of late sudden graft failure. There is no previous report analysing survival up to ten years after CABG in patients with different degree of LMCA obstruction.

The purpose of this study was to compare patient characteristics and to analyze deaths up to ten years after CABG in patients without LMCA stenosis, with low-grade stenosis, with high-grade stenosis and patients with chronic total occlusion of the LMCA.

Patients and methods

During 1970–1989 totally 3 866 patients had isolated primary CABG at the Karolinska Hospital in Stockholm, Sweden, 496 of whom (13%) had significant LMCA obstruction. During this period we encountered 11 patients with an acquired chronic total occlusion of the LMCA. Up to 2002, four additional patients with this rare lesion were operated on at this centre up corresponding to 0.1% (15/12_142) of all patients undergoing isolated CABG during 1970–2002.

Operative technique

The operations in this report were performed via median sternotomy with the aid of cardiopulmonary bypass, haemodilution (haematocrit 20–30%) and moderate hypothermia (25–30°C). Membrane oxygenators completely replaced previously used disposable bubble oxygenators during the 1980s. In most patients myocardial protection was achieved by antegrade hypothermic, crystalloid cardioplegia. The distal anastomoses were constructed with a continuous running 7-0 polypropylene suture and the proximal vein anastomoses to the aorta sutured over a partially occluding clamp while reperfusing the heart and rewarming the patient.

Angiographic evaluation

According to the medical records 552 patients were operated on because of LMCA disease. The angiographies could be retrieved and were reviewed for 408 patients (74%) by two thoracic surgeons (AJ and JL) together until consensus was reached. For 144 patients the original detailed angiographic report was again scrutinized. The examiners agreed on the decision that the LMCA was not significantly obstructed in 56 patients (10%) leaving 496 patients with a significant LMCA obstruction.

Any coronary artery stenosis was considered haemodynamically significant if the luminal diameter was reduced at least 50% in two projections. The LMCA was classified by visual analysis of the angiography from at least two projections as either without stenosis, with low-grade stenosis if there was more than 50% but less than 75% narrowing, with high grade stenosis if the narrowing was more than 75% or completely occluded if no lumen was filled when an injection of contrast was performed into the ostium of the left coronary artery. The right coronary artery was dominant in patients with occlusion of the LMCA with extensive collaterals supplying the left ventricle via intercoronary collateral vessel patterns including atrial branches as the sinus node and atrioventricular artery, the conus artery, septal and apical branches as described by others 8, 9 (Figure 1).

Figure 1.  Contrast injected into the right coronary artery in a patient with complete total occlusion of the left main coronary artery. The left ventricular epicardial arteries are filled via extensive collaterals.

Definitions and statistical methods

Patient data from operations performed at the Karolinska hospital during 1970–2002 were continuously recorded in the department database and were extracted by retrospective review of all medical records. The study was approved by the hospital's Ethics Committee and informed consent was obtained to include data in the database. Demographic variables considered in this study included age, gender, and year of surgery. Health indicator variables included history of hypertension, hyperlipidaemia, diabetes, previous acute myocardial infarction, unstable angina, as well as, serum creatinine. Creatinine clearance was calculated using the equation of Cockroft and Gault 10.

Left ventricular function was classified based on physician assessment of angiography or echocardiography as reported in the medical record. If there were dyskinetic left ventricular wall segments and an estimated ejection fraction better than 30% and less than 50%, the function was classified as reduced. A severely reduced left ventricular function was defined as a markedly dilated ventricle with one or more akinetic or hypokinetic segments with an estimated ejection fraction less than 30%.

Follow-up of mortality was performed in all patients by record linkage to the Swedish National Cause of Death Register for the period 1970–2000 using the Swedish personal identification number. This register is complete for residents of Sweden and there was no loss of deaths during the follow-up and no emigration out of the country in this cohort of patients. All patients were followed until death or up to December 31^st 2000 whichever came first with regard to vital status. Because of the few patients with chronic total occlusion of the LMCA we added four patients with this lesion operated on after 1989 and followed this group up to September 2005. Early mortality was defined as death within 30 days of the operation. Kaplan-Meier estimates of survival up to ten years after surgery were calculated.

Early mortality was estimated for patients with different degree of LMCA stenosis and comparisons were performed crude and adjusted by odds ratios (OR) obtained from logistic regression analyses. Proportional hazards regression was used to estimate crude and adjusted relative risks (RR) for long term mortality in the different patient groups compared to patients with no LMCA stenosis. In these analyses random variation was taken into account by calculating 95% confidence intervals (CI). Because of a small total number of patients (n=15) with chronic total occlusion of the LMCA, the observed deaths within five and ten years after surgery, respectively, was compared with the expected number of deaths based on the mortality in all patients without LMCA stenosis by calculation of standardized mortality ratios. In the estimation of expected numbers consideration was taken to gender and age (≤ 60 years and > 60 years). The 95% CI for the standardized mortality ratio was estimated using exact methods and assuming that the number of observed deaths followed a Poisson distribution. Continuous variables are presented as means with one standard deviation and differences in means between two patient groups were evaluated by Student's t-test. The χ² test or the Fischer's exact test was used to analyze proportions for categorical variables.

Results

Patient characteristics in relation to degree of LMCA obstruction

Characteristics of all patients undergoing a first isolated CABG without significant LMCA stenosis, those with high-grade stenosis, those with low-grade stenosis and with chronic total occlusion of the LMCA are presented in Table I. Patients with a high-grade LMCA stenosis were older (p < 0.001) and peripheral vascular disease was more common (p < 0.001) than in patients without LMCA stenosis. Older age (p = 0.05), hyperlipidemia (p < 0.001) and unstable angina (p = 0.05) was more common in patients with chronic occlusion of the LMCA than in patients without LMCA stenosis. Reduction of left ventricular function did not correlate to the degree of LMCA obstruction.

Table I.  Characteristics of patients undergoing coronary artery bypass in relation to degree of left main coronary artery obstruction.

		Left main stem pathology
	No stenosis (n = 3370)	Low-grade stenosis (n = 261)	High-grade stenosis (n = 224)	Complete occlusion (n = 15)
Age (years)	58±8	60±8	61±8	62±8
Body mass index (kg/m^2)	26±4	26±3	26±4	27±4
Creatinine clearance (ml/min)	83±23	83±25	80±25	78±18
No. distal anastomoses	2.9±3.7	3.6±1.1	3.4±1.2	3.1±0.7
Mammary artery graft (%)	(74)	(77)	(72)	(87)
Female gender (%)	(15)	(13)	(19)	(13)
Hypertension (%)	(26)	(26)	(28)	(40)
Hyperlipidemia (%)	(8)	(10)	(6)	(40)
Diabetes mellitus	(9)	(10)	(8)	(13)
Previous AMI (%)	(54)	(52)	(35)	(53)
Unstable angina (%)	(6)	(8)	(8)	(20)
History of stroke (%)	(3)	(7)	(5)	(0)
Peripheral vascular disease (%)	(6)	(6)	(12)	(7)
Year of surgery
1970–1979 (%)	(18)	(6)	(12)	(20)
1980–1989 (%)	(82)	(94)	(88)	(53)
No. obstructed arteries
One vessel (%)	(16%)
Two vessel (%)	(34%)	(24%)	(23%)	(33%)
Three vessel (%)	(47%)	(73%)	(69%)	(67%)
Missing (%)	(1%)	(3%)	(8%)
Left ventricular function
Normal (%)	(58)	(59)	(61)	(60)
Reduced (%)	(35)	(37)	(32)	(27)
Severely reduced (%)	(4)	(3)	(4)	(13)
Missing (%)	(3)	(1)	(3)	(0)

AMI = acute myocardial infarction.

Mortality

Survival up to ten years after the operation in patients without LMCA stenosis, in those with patients with low-grade and high-grade stenosis, respectively, and in patients with total occlusion of the LMCA is depicted in Figure 2. Survival at ten years was 76% (95% CI 74–77) in patients without LMCA stenosis, 74% (95% CI 68–79) in patients with low-grade stenosis and 64% (95% CI 58–70) in patients with high-grade stenosis of the LMCA. The late deaths occurring in patients with total occlusion of the LMCA resulted in a 10-year survival of 62% (95% CI 31–82).

Figure 2.  Survival in patients without stenosis, low-grade stenosis, high-grade stenosis and chronic total occlusion of the left main coronary artery. Number of remaining patients at five and ten years are indicated.

The crude early death rate was three times higher in patients with high-grade LMCA stenosis (6.3%) than in patients without LMCA stenosis (1.9%) (Table II). This high early mortality remained after multivariate adjustment (OR 2.6, 95% CI 1.4–4.8). Low-grade LMCA stenosis was not associated with an increased risk of early death. There was no early mortality in the patients with chronic total occlusion of the LMCA.

Table II.  Early, 5-year and 10-year mortality after coronary artery bypass grafting in relation to degree of left main coronary artery stenosis.

Time period	Left main coronary artery pathology
	No stenosis n = 3370	Low-grad stenosis n = 261		High-grade stenosis n = 224
≤30 days Deaths	63 deaths (1.9%)	6 deaths (2.3%)		14 deaths (6.3%)
	OR	OR	95% CI	OR	95% CI
Crude	1.0	1.2	0.5–2.9	3.5	1.9–6.4
Adjusted for age and year of surgery	1.0	1.1	0.5–2.5	2.7	1.5–5.0
Multivariate adjustment ^a	1.0	1.0	0.4–2.4	2.6	1.4–4.8
>30 days and ≤5 years Deaths	n = 3307 297 deaths (9.0%)	n = 255 25 deaths (9.8%)		n = 210 29 deaths (13.8%)
	RR	RR	95% CI	RR	95% CI
Crude	1.0	1.1	0.7–1.7	1.6	1.1–2.3
Adjusted for age	1.0	1.0	0.7–1.5	1.4	0.9–2.1
Multivariate adjustment ^b	1.0	0.9	0.7–1.2	1.4	1.1–1.8
> 30 days and ≤10 years Deaths	n = 3307 753 deaths (22.7%)	n = 255 62 deaths (24.3%)		n = 210 66 deaths (31.4%)
	RR	RR	95% CI	RR	95% CI
Crude	1.0	1.1	0.8–1.4	1.5	1.2–1.9
Adjusted for age	1.0	1.0	0.7–1.3	1.3	1.0–1.7
Multivariate adjustment ^c	1.0	1.0	0.8–1.3	1.5	1.1–1.8

^a Adjusted for age, year of surgery, previous acute myocardial infarction, number distal anastomoses and number obstructed coronary vessels.

^b Adjusted for age, previous acute myocardial infarction and number obstructed coronary vessels.

^c Adjusted for age and previous acute myocardial infarction.

OR = odds ratio; CI = confidence interval; RR = relative risk.

Exclusive of early deaths there was a 40% and 50% increased mortality, respectively, after multivariate adjustment at five and ten years of follow-up in patients with high-grade stenosis compared to patients without LMCA stenosis (Table II). No increased risk of late death was observed in patients with low-grade LMCA stenosis. During the initial five years after surgery one patient with total occlusion of the LMCA died. Based on the experience of patients without LMCA stenosis 1.9 deaths were expected taking age and gender into account yielding a standardized mortality ratio of 0.5 (95% CI 0.0–3.0). At ten years the observed total number of five deaths in patient with total occlusion of the LMCA was somewhat higher but still close to the calculated 3.8 corresponding to a standardized mortality ratio of 1.3 (95% CI 0.4–3.0).

Ischaemic heart disease caused 62% of the late deaths (547/886 deaths). All deaths in patients with total occlusion of the LMCA were cardiac. Death from ischaemic heart disease tended to be more common (p = 0.09) in patients with increasing severity of LMCA obstruction (Table III).

Table III.  Causes of late deaths during two five-year periods after coronary artery bypass grafting in relation to degree of left main coronary artery obstruction.

Time period and cause of death	Left main coronary artery pathology
> 30 days ≤ 5 years	No stenosis (n = 3307)	Low-grade stenosis (n = 255)	High-grade stenosis (n = 210)	Complete occlusion (n = 15)
Total deaths	n = 297	n = 25	n = 29	n = 1
Ischaemic heart disease (%)	185 (5.6)	18 (7.1)	18 (8.6)	1 (9)
Stroke (%)	22 (0.7)	3 (1.2)	2 (1)	0
Malignancy (%)	51 (1.5)	1 (0.4)	5 (2.4)	0
Other causes (%)	39 (1.2)	3 (1.2)	4 (1.9)	0
>5 ≤10 years	n = 3010	n = 230	n = 181	n = 14
Total deaths	n = 456	n = 37	n = 37	n = 4
Ischaemic heart disease (%)	278 (9.2)	23(10.0)	22 (12.2)	4 (29)
Stroke (%)	25 (0.8)	3 (1.3)	3 (1.7)	0
Malignancy (%)	66 (2.2)	5 (2.2)	5 (2.8)	0
Other causes (%)	87 (2.9)	6 (2.6)	7 (3.9)	0

Discussion

In this report increasing severity of LMCA stenosis seemed to increase the risk of early mortality after CABG. High-grade LMCA stenosis in contrast to low-grade LMCA stenosis was an independent risk factor for early mortality both crude and after correction for differences in age, year of surgery and other confounding factors. According to experience from the Cleveland Clinic Foundation left main disease was completely neutralized as a risk factor for operative mortality in the 1980s but in that study the degree of LMCA obstruction is not assessed 11. Our study is unique because of the long follow-up of patients with different degree of LMCA obstruction. We have only found two reports attempting to correlate the degree of LMCA obstruction to the outcome after CABG but none followed the patients longer than 3.5 years 12, 13. In agreement with our findings of an increased early mortality in patients with high-grade LMCA stenosis, Kennedy and associates have documented an operative mortality of 1.6% in patients with mild LMCA stenosis, 3.6% if the LMCA stenosis was 75–89% and 7.9% if the LMCA was stenosed ≥90% 12. Despite the alarming angiographic finding of an occluded LMCA and the associated jeopardized myocardium the operations were performed without early mortality in consistency with the experience in other reports 14–16.

High-grade but not low-grade LMCA stenosis was an independent risk factor for late death. Our study showed had 40–50% higher late mortality in patients with a high-grade LMCA stenosis than in those without LMCA obstruction. Takaro and coworkers conclude that LMCA disease is not a homogenous disease 13. They reported survival of 83% at 42 months after CABG in patients with > 75% LMCA stenosis and 92% in those with 50–75% LMCA stenosis in a group of 48 patients randomized to surgical treatment. In view of the small number of patients with total occlusion of the LMCA the risk analysis of late death in our series must be interpreted with caution. We found no evidence that late mortality differed substantially after CABG in patients with total occlusion of the LMCA and in those without LMCA obstruction.

Significant LMCA obstruction was present in 13% of the reported patients who had CABG up to 1989 and in our experience about one of 1 000 patients undergoing CABG had chronic total occlusion of the LMCA. Older age and peripheral vascular disease in one tenth of patients with a high-grade stenosis of the LMCA probably reflected the severity of the atherosclerotic disease. We identified a small group of patients with the capability to develop sufficient collaterals from a dominant right coronary artery to allow survival even though the LMCA was completely occluded. We were not able to define specific clinical markers to distinguish these patients with LMCA occlusion from other patients with obstructive coronary artery disease.

In our series about 60% of patients with significant LMCA obstruction had normal left ventricular. Most patients with a significant left main stenosis appear to have a preserved left ventricular ejection fraction, normal wall motion, and no significant alteration of the left ventricular end-diastolic pressure function at rest reflecting the significance of collateral blood supply 17. Collateral circulation from the right coronary artery via septal arteries and other branches that provide the entire myocardial blood supply to the left ventricle and have been documented in the extreme subgroup of patients with a chronic occluded LMCA 14, 15. This extensive collateral circulation may preserve ventricular function but is invariably insufficient to prevent myocardial ischaemia during stress 14, 18. The presence of poor collateral circulation and associated obstruction of the RCA will impair prognosis 7. A bypass operation in patients with a high-grade coronary artery obstruction may dramatically change collateral circulation pattern making these patients vulnerable to events such as graft failure a possible mechanism for late occurring deaths 19. Furthermore the extent of collaterals decreases with age 20. The increased long-term mortality in patients with high-grade LMCA stenosis could at least in part be due to regression of collaterals after the operation and events occurring after sudden graft failure.

There are limitations of our study. The degree of LMCA stenosis could not be precisely graded only by scrutinizing angiographies or reports. We feel confident that patients classified as having a high-grade LMCA stenosis definitely had worse obstruction than those classified as having a low-grade LMCA stenosis and there were certainly no misclassifications between the groups with no stenosis and high-grade stenosis. Any misclassification of grade of stenosis is unlikely to explain the finding that early and long-term mortality was higher in patients with high-grade LMCA stenosis than in those without LMCA stenosis. Rather, this difference in mortality was most likely real and may reflect more severe atherosclerotic disease in patients with high-grade stenosis, difficulties to obtain optimal myocardial protection during the operation, dependence upon function of grafts and the imponderable regression or persistence of coronary collaterals in case of late occurring bypass graft failures. Unfortunately we did not have autopsy data to correlate late deaths to patency of grafts. A definite strength of this study is the complete follow-up of mortality during ten years due to reliable population registers and the national cause of death register without loss of deaths. In spite of only 15 patients with chronic total occlusion of the LMCA there is no larger series reported from one institution.

We conclude that high-grade LMCA stenosis correlated to a higher risk of early and late death after CABG than in patients without LMCA obstruction indicating that these patients require particular attention during and after the operation. Patients with a low-grade LMCA stenosis had a prognosis after CABG similar to that in patients without LMCA obstruction. There was substantially no increased long-term mortality in the few patients with occlusion of the LMCA.

The study was supported by a grant from Stockholm County Council (EXPO 95). We are grateful to Christian Unge, Henrik Overödder, and Björn Törnkvist for the abstraction of medical record data.