Does Reasonable Incomplete Surgical Revascularization Affect Early or Long-Term Survival in Patients With Multivessel Coronary Artery Disease Receiving Left Internal Mammary Artery Bypass to Left Anterior Descending Artery?
Background— The objective was to evaluate the impact of complete revascularization (CR) versus reasonable incomplete surgical revascularization (IR) in others than left anterior descending artery territory on early and late survival in patients with multivessel coronary artery disease (CAD).
Methods and Results— During a 7-year period, 8.806 consecutive patients with multivessel CAD affecting the proximal left anterior descending artery or left main stem underwent sternotomy for isolated coronary artery bypass grafting including left internal mammary artery-left anterior descending artery bypass. A total of 936 patients (10.6%) had IR of the circumflex or right coronary artery territory. IR was based on the traditional classification. Follow-up was 3.5±2.2 years. Patient groups were comparable regarding age (CR 67.1 versus IR 67.6 years), ejection fraction (57.2% versus 57.5%), and logEuroscore (4.5% versus 4.5%). Patients receiving IR presented with more complex CAD. Mean number of distal anastomoses was 3.0±0.8 CR versus 2.4±0.6 IR. Operation time (176 versus 187 minutes) and cross-clamp time (52 versus 56 minutes) were longer in the IR group (P<0.001). Hospital mortality was 3.3% CR versus 3.2% IR (P=0.520). Independent risk factors for hospital mortality were age, pulmonary hypertension, chronic obstructive pulmonary disease, peripheral vascular disease, and emergency indication, but not IR (P=0.922). Arterial revascularization was protective. Cumulative survival at 1-year was 93.1% CR versus 93.6% IR and at 5 years 82.2 CR versus 80.9% IR (P=0.457).
Conclusions— In presence of left internal mammary artery-to-left anterior descending artery bypass, reasonable IR of the circumflex or right coronary artery territory did not adversely affect early or long-term survival in patients with multivessel CAD. In patients presenting with 1 poor target vessel; however, IR is a good therapeutic option and the benefit of CR should be balanced against the risks.
Complete revascularization (CR) has been considered the standard approach in coronary artery disease (CAD). This has partially been based on a study from 1981, reporting a 5-year survival rate of 84% in patients with CR compared to 69% in those with incomplete revascularization (IR).1 Nevertheless, data on this dogma remain very limited so far and therefore this issue is not precisely defined and highlighted in the contemporary guidelines for coronary artery bypass graft (CABG) surgery.2 Moreover, studies validating this principle were performed in the early CABG era on younger patient populations with predominantly saphenous vein graft use and included patients with calculated IR of potentially good graftable vessels based on contemporary interpretation.
However, even when CR is a major goal, there remain several reasons to perform reasonable surgical IR, based on preoperative assessment (eg, nondominant diseased right coronary artery, non-vital myocardium, attempt to avoid cardiac arrest for off-pump coronary artery bypass in high-risk patients, limited graft material), or more unexpected findings during the operation (ie, small target vessel, severely calcified artery, vessel not to be found, coronary injury).
Different definitions of CR exist in the literature. First, the traditional definition assumes CR when at least 1 bypass is performed in a significantly diseased primary arterial territory and is predominantly used in the early CABG era, including during the CASS study.3–7 Second, the functional classification asks for bypass grafting to all major diseased and graftable coronary artery segments with significant flow-compromising lesions and was favorably approached in comparative randomized CABG versus percutaneous coronary intervention studies.8,9 Third, the index for completeness of revascularization, the surgical intention-to-treat ratio of the number of intraoperatively performed divided by the number of preoperatively planned distal coronary anastomoses can be used; it should be ≥1 for CR.10,11 These varying definitions and significant interobserver and interinstitutional variability regarding diseased and graftable coronary artery vessels seriously interfere with comparative data analysis on incomplete revascularization.
It is well known that in CAD, left main and proximal left anterior descending artery (LAD) lesions have the highest prognostic relevance.12,13 Thus, it can be speculated that bypass grafting of the LAD with the left internal mammary artery (LIMA) may mainly compensate the prognostic impact of multivessel CAD. CABG surgery has evolved in the past decade, allowing an individualized surgical concept including total arterial bypass grafting and off-pump coronary artery bypass surgery. To provide concurrent clinical data, we analyzed the early and late outcome of patients presenting with prognostic CAD including proximal LAD or left main stem lesions receiving LIMA-to-LAD grafting plus additional CR or reasonable IR.
Patients and Methods
From January 2000 until August 2007, a total of 11 269 consecutive patients underwent isolated CABG with LIMA-to-LAD bypass with proximal LAD or left main stem disease. Of these, 1145 patients received minimally invasive direct coronary artery bypass (MIDCAB) procedure, 373 were in deep cardiogenic shock, 474 had redo CABG, and 471 received sternotomy for 1-vessel disease of the LAD and were excluded from the study. A total of 8806 patients received sternotomy for prognostic relevant coronary multivessel disease as a primary CABG procedure and formed the study population. In patients after previous percutaneous coronary intervention, only coronary segments with target vessel restenosis or occlusion were considered as diseased. The traditional classification of IR was applied; 936 patients (10.6%) received IR because of small-vessel diameter in 63.1%, because of severely calcified target vessels in 25.3%, because of myocardial scar in 3.9%, and because of other reasons in 7.7%. Among the IR patients, 250 (26.7%) received no bypass graft to the circumflex and 686 (73.7%) received no bypass graft to the right coronary artery territory.
Diabetes was defined as hyperglycemia requiring medical treatment. Renal insufficiency and pulmonary hypertension was defined according to the EuroSCORE variables.14
Onset of postoperative acute myocardial infarction was defined as type 5 (associated with CABG), according to the Consensus Document for universal definition of myocardial infarction.15 Postoperative acute renal failure was assumed as requirement for continuous venovenous hemofiltration. Stroke was defined as transient or persisting postoperative hemiparesis or neurological dysfunction with morphological substrate confirmed by computer tomography or nuclear magnetic resonance. Major adverse cardiac and cerebrovascular events were assessed for all major cerebrovascular and cardiac events, including death of all cause.
Follow-up was 3.5±2.2 years. It was 100% complete for vital status and censored either at the time of last available follow-up or when death occurred. Follow-up was performed using a stepwise protocol of routine postal questionnaire, written or telephone contact to the patient, telephone contact to the family doctor, and contact at least to the local governmental office.
Full surgical revascularization was the goal in all patients preoperatively and hybrid procedures were not primarily scheduled. Routine sternotomy was performed and internal mammary artery takedown was performed using a skeletonized technique. LIMA was used in all patients for revascularization of the LAD. Cardiopulmonary bypass was established by standard ascending aortic and right atrial cannulation. Operative details of off-pump coronary artery bypass and conventional bypass surgery have been described previously.16
All data were prospectively recorded. Continuous variables are expressed as mean±SD and categorical data are expressed as proportions throughout the article. Categorical variables were compared using the χ2 test and independent continuous variables were compared by 2-tailed Student test or Mann-Whitney U test as appropriate at a level of significance of P<0.05. Univariate and multivariate logistic regression model with stepwise backward procedure was performed to assess the impact of incomplete revascularization including 26 preoperative and 20 operative binary variables on hospital and cumulative survival (Tables 1 and 2⇓). Results are described as odds ratio (OR) and 95% confidence interval (CI). Survival was calculated by Kaplan-Meier methods and log-rank test. The statistical analyses were performed using 14.0 SPSS software package.
Statement of Responsibility
All authors had full access to the data and take full responsibility for its integrity. They read and agree to the manuscript as written. The study was approved by the local Ethics Committee and study design, anonymous data acquisition, as well as data publication were approved according to the declaration of Helsinki.
Patient characteristics are presented in Table 1. For patients receiving CR versus IR, significant differences were found regarding the prevalence of diabetes, hyperlipidemia, arterial and pulmonary hypertension, and complexity of CAD. Patients with emergency surgical indication had a lower rate of IR. There was also a strong trend for more IR in patients with previous percutaneous coronary intervention treatment (P=0.058). Significant risk factors for IR by univariate logistic analysis expressed as OR are given in Table 1 and depicted in Figure 1. Of these, 3-vessel CAD (OR, 1.56; P=0.002) was a significant risk factor for incomplete revascularization by multivariate analysis, whereas 2-vessel CAD (OR, 0.66; P=0.003), emergency indication (OR, 0.63; P=0.046), left main disease (OR, 0.69; P=0.004), and off-pump coronary artery bypass (OR, 0.343; P=0.001) were protective.
Procedural details and operative data are demonstrated in Table 2. Despite a significantly lower number of distal coronary anastomoses (CR 2.97 versus IR 2.44 per patient), procedural times were slightly but significantly longer in patients with IR. In off-pump coronary artery bypass surgery IR was less common than in conventional CABG procedures. Patients receiving IR had a significantly higher rate of total arterial revascularization but less sequential grafting. The difference in bypass grafting to the right coronary artery territory (CR 67.1 versus IR 44.6%) was more obvious than to the circumflex territory (CR 87.0 versus IR 73.6%).
There were no major differences in postoperative complications, including major adverse cardiac and cerebrovascular event rate between CR and IR patients (Table 3). Hospital mortality was 3.3% for CR and 3.2% for IR (P=0.520). Univariate analysis for hospital mortality of 46 preoperative and intraoperative variables revealed several significant risk factors for hospital mortality, whereas arterial grafting was protective (Figure 2). Of all the potential risk factors, only age older than 70 years (OR, 2.23; P<0.001), chronic obstructive pulmonary disease (OR, 1.79; P=0.010), peripheral vascular disease (OR, 1.51; P=0.029), renal insufficiency (OR, 3.13; P<0.001), pulmonary hypertension (OR, 2.58; P=0.013), emergency indication (OR, 1.70; P<0.028), high inotropic support (OR, 1.52; P<0.001), and hyperlipidemia (OR, 0.60; P=0.004) were found to be independent predictors for hospital mortality by multivariate analysis. Most importantly, IR did not significantly affect hospital mortality (OR, 0.98; P=0.922).
Overall cumulative survival was comparable between CR and IR patients (Figure 3). The 1-, 3-, and 5-year survival rates were 93.0±0.3%, 88.3±0.4%, and 82.2±0.5% for CR compared to 93.6±0.8%, 87.0±1.3%, and 80.9±1.8% for IR (P=0.457).
Subgroup analysis demonstrated no significant survival difference in patients with high extracardiac morbidity (logEuroSCORE >6). For these patients 5-year survival was 61.3±2.5% for complete and 54.0±3.2% for incomplete revascularization (P=0.768). In addition, IR did not adversely affect the survival in the subset of patients with 3-vessel CAD or left main stem stenosis (data not shown). As demonstrated in Figure 4, the primary territory of incomplete revascularization did not have an influence on cumulative survival. After 1 and 5 years, cumulative survival rates were 92.0±2.2% and 88.8±2.8% for IR of the circumflex territory versus 95.0±0.9% and 81.0±2.9% for IR of the right coronary artery territory, respectively (P=0.718).
In multivariate analysis, independent risk factors for overall mortality were age older than 70 years (OR, 1.52; P<0.001), chronic obstructive pulmonary disease (OR, 1.84; P=0.010), peripheral vascular disease (OR, 1.52; P<0.001), renal insufficiency (OR, 3.64; P<0.001), diabetes (OR, 1.23; P=0.033), atrial fibrillation (OR, 1.78; P=0.002) and logEuroSCORE >6 (OR, 1.66; P<0.001), whereas hyperlipidemia (OR, 0.77; P=0.012) and bilateral mammary grafting (OR, 0.56; P<0.001) were associated with a significantly better outcome. However, IR did not affect overall survival (OR, 0.89; P=0.234).
CR is a major goal in CABG based on the long-existing principle that CR is superior to IR regrading long-term survival to early and long-term survival. However, reasonable IR occurs in the clinical practice but varies significantly between 9% to 39% in the literature, which is, among others, according to the non-standardized definition of IR.3–8,11,17,18 Nevertheless, it might be crucial for some CV surgeons to admit having performed IR and to deal with its consequences, although reasons for IR were associated with patient-specific coronary pathology rather than the surgical quality. Prognostic impact of IR strongly varies in the limited number of retrospective analyses.3–8,11,17–19 This is also because of the use of different IR definitions, but it is also a result of different CABG era reports, including LIMA use, different patient populations, inhomogeneous CAD pattern, and varying follow-up periods. We here focused on the prognostic impact of IR of circumflex and right coronary artery territory in patients with lesions of the most important coronary segments, ie, the proximal LAD or left main stem. When performing LIMA-to-LAD bypass, it was hypothesized that reasonable IR of the circumflex or right coronary artery territory does not adversely affect survival. However, it must be stressed that reasonable surgical IR is strongly different from IR in percutaneous coronary intervention of multivessel disease when frequently good graftable vessels with large areas of viable myocardium remain untreated.20
One of the major findings of the study was that hospital and late survival were comparable for IR and CR patients in presence of LIMA-to-LAD grafting. The issue of prognostic consequences of IR versus CR was highlighted by 3 reports3–5 from the early CABG era using the traditional definition of IR. In a CASS study report of 3372 CABG patients with 3-vessel CAD who underwent operation between 1974 and 1979, no significant difference in operative mortality was found according to the number of grafted vessels. However, follow-up mortality showed a strong association with the number of performed bypasses, especially in the event of single bypass grafting. Jones et al4 found similar consequences of IR regarding hospital and follow-up mortality in 2857 patients with multivessel CAD who underwent operation between 1978 and 1981 and during a 12-year follow-up. They also found that the number of diseased vessels and the left ventricular ejection fraction were the strongest predictors for IR. In a Cleveland Clinic analysis of 2067 patients treated from 1971 to 1997 with isolated LIMA-to-LAD bypass, those with additional nonrevascularized disease of the circumflex or right coronary artery territory demonstrated a significantly reduced long-term survival.5
However, in contrast to them, we found no impact of IR on long-term survival. One reason for this discrepancy might be based on differences between contemporary patient populations and those of the early CABG period regarding age, life expectancy, cardiac and extracardiac comorbidities, and the infrequent use of internal thoracic artery grafting.3,4 Additionally, it might be reasonable that progression in modern secondary pharmacological treatment could compensate surgical IR in diffuse distal CAD of the non-LAD territories. It also has to be addressed that the follow-up was limited in our analysis compared to the aforementioned studies reporting extensive follow-up up to 20 years. This might be of relevance because survival curves between CR and IR have the potential to continually separate over time.3–5
During the BARI study from 1988 to 1991, the investigators were unable to demonstrate a survival advantage of CR in 1526 patients with >80% internal mammary artery use.19 In contrast Kleisli et al17 reported a significant benefit in long-term survival for CR patients using contemporary strategies of arterial grafting. Other studies from the routine LIMA-to-LAD grafting era focused on IR in a limited number of selected patients and were conflicting regarding early and late outcome.6–8,10,11,17,18 In 1479 off-pump coronary artery bypass patients, Caputo et al reported a reduced survival in IR patients, which was evident only during the first 6 months after surgery.6 They speculated that the preoperative condition of the patients rather than IR itself was the reason for this finding. Comparably, IR increased the early mortality in 263 patients with impaired left ventricular function and in 859 patients older than 75 years of age.10,11 In 500 octogenarians, Kozower et al7 reported reduced hospital and long-term survival for IR, whereas Moon et al18 found CR not to be associated with improved survival in octogenarians.7,18 It can be supposed that a higher risk of death early after IR might be simply a surrogate for advanced CAD and more extracardiac morbidities.11 This is partially supported by our study results, because patients with IR presented more often with diabetes, hyperlipidemia, and arterial hypertension. However, compared to others, this was not translated into higher perioperative mortality in our series.
As in almost all CABG studies, we used the traditional classification for CR, allowing a reliable data recording.3–7,17,18 Comparing both traditional and functional classification in 1507 CABG patients from the BARI study, the investigators found no independent long-term survival difference for traditional or functional classification regarding CR or IR.19 Interestingly, they revealed the highest adverse event rate in patients receiving >1 distal anastomosis to a non-LAD territory. An early adverse effect of overly excessive surgical revascularization on major perioperative complications and acute myocardial infarction was also described recently and might be partially explained by a 25% venous graft failure rate during 1-year angiographic follow-up.21,22
Not surprisingly we identified several patient-specific risk factors for cumulative mortality. Of utmost importance was the finding that arterial bypass grafting with bilateral mammary artery use improved survival significantly and independently of the completeness of revascularization. It might be speculative that the lower number of distal anastomoses might be compensated by the higher number of arterial grafts in IR patients. However, bilateral mammary artery use is protective and should be considered in all CABG patients. We also found hyperlipidemia to improve long-term survival. It can be speculated that this might be a surrogate for optimized lipid-lowering medication, its positive effects on cardiovascular event rate, and protective pleiotropic effects.23
There are some limitations of the present study. First, it was subject to all limitations inherent to a nonrandomized study, including potential selection bias regarding which patients underwent CR versus IR. This was addressed by identifying selection criteria for IR and using a multivariate logistic regression model to identify independent risk factors. However, such biases are impossible to fully eliminate when patient analysis is retrospective. Second, during follow-up we only focused on long-term survival. This is because complete follow-up data recording of angina status, major cardiac events, and repeat revascularization was possible in only 82% during the study period, and further efforts have to be made to complete these follow-up data. We also cannot fully exclude that the difference between IR and CR becomes significant over time.
The present study was performed to analyze the outcome of CABG patients who received IR versus CR in others than LAD territories. Small distal or severely calcified target vessels were the most frequent reasons for IR. Slightly >10% of all patients treated during a 7-year period received IR. The important finding of this study was that this subgroup of patients performed reasonably well. One- and 5-year mortality rates did not differ between patients receiving CR versus IR. Despite this positive finding, surgeons should not be misled to perform IR in patients with graftable vessels.
Because of modern armamentarium, surgeons now more so than in former decades can adapt their surgical strategy based on the individual patient risks. Bilateral mammary artery use was protective. For patients presenting with 1 poor target vessel, however, IR is a good therapeutic option and the benefit of CR should be balanced against the risk. Thus, hybrid concepts with LIMA-to-LAD via lateral thoracotomy (MIDCAB) might be an alternative to sternotomy in patients with significant surgical risk for IR. These data might also support an individual decision of less complete off-pump coronary artery bypass revascularization in patients at high risk for cardiopulmonary bypass.
Presented in part at American Heart Association Scientific Sessions 2008, November 8–12, 2008, New Orleans, La.
Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, Gardner TJ, Hart JC, Herrmann HC, Hillis LD, Hutter AM Jr, Lytle BW, Marlow RA, Nugent WC, Orszulak TA. American College of Cardiology; American Heart Association. ACC/AHA 2004 guideline update for coronary artery bypass graft surgery: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1999 Guidelines for Coronary Artery Bypass Graft Surgery). Circulation. 2004; 110: e340–e437.
Bell MR, Gersh BJ, Schaff HV, Holmes DR Jr, Fisher LD, Alderman EL, Myers WO, Parsons LS, Reeder GS. Effect of completeness of revascularization on long-term outcome of patients with three-vessel disease undergoing coronary artery bypass surgery. A report from the Coronary Artery Surgery Study (CASS) Registry. Circulation. 1992; 86: 446–457.
Scott R, Blackstone EH, McCarthy PM, Lytle BW, Loop FD, White JA, Cosgrove DM. Isolated bypass grafting of the left internal thoracic artery to the left anterior descending coronary artery: late consequences of incomplete revascularization. J Thorac Cardiovasc Surg. 2000; 120: 173–184.
van den Brand MJ, Rensing BJ, Morel MA, Foley DP, de Valk V, Breeman A, Suryapranata H, Haalebos MM, Wijns W, Wellens F, Balcon R, Magee P, Ribeiro E, Buffolo E, Unger F, Serruys PW. The effect of completeness of revascularization on event-free survival at one year in the ARTS trial. J Am Coll Cardiol. 2002; 39: 559–564.
Ong AT, Serruys PW, Mohr FW, Morice MC, Kappetein AP, Holmes DR Jr, Mack MJ, van den Brand M, Morel MA, van Es GA, Kleijne J, Koglin J, Russell ME. The SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX) study: design, rationale, and run-in phase. Am Heart J. 2006; 151: 1194–1204.
Osswald BR, Blackstone EH, Tochtermann U, Schweiger P, Thomas G, Vahl CF, Hagl S. Does the completeness of revascularization affect early survival after coronary artery bypass grafting in elderly patients? Eur J Cardiothorac Surg. 2001; 20: 120–125.
Yusuf S, Zucker D, Peduzzi P, Fisher LD, Takaro T, Kennedy JW, Davis K, Killip T, Passamani E, Norris R. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration. Lancet. 1994; 344: 563–570.
Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999; 16: 9–13.
Thygesen K, Alpert JS, White HD. Joint ESC/ACCF/AHA/WHF Task Force for the Redefinition of Myocardial Infarction, Universal definition of myocardial infarction. Circulation. 2007; 116: 2634–2653.
Rastan AJ, Eckenstein JI, Hentschel B, Funkat AK, Gummert JF, Doll N, Walther T, Falk V, Mohr FW. Emergency coronary artery bypass graft surgery for acute coronary syndrome: beating heart versus conventional cardioplegic cardiac arrest strategies. Circulation. 2006; 114 (1 Suppl): I477–I485.
Hannan EL, Racz M, Holmes DR, King SB III, Walford G, Ambrose JA, Sharma S, Katz S, Clark LT, Jones RH. Impact of completeness of percutaneous coronary intervention revascularization on long-term outcomes in the stent era. Circulation. 2006; 113: 2406–2412.
Alamanni F, Dainese L, Naliato M, Gregu S, Agrifoglio M, Polvani GL, Biglioli P, Parolari A. Monzino OPCAB Investigators. On- and off-pump coronary surgery and perioperative myocardial infarction: an issue between incomplete and extensive revascularization. Eur J Cardiothorac Surg. 2008; 34: 118–126.
Magee MJ, Alexander JH, Hafley G, Ferguson TB Jr., Gibson CM, Harrington RA, Peterson ED, Califf RM, Kouchoukos NT, Herbert MA, Mack MJ. PREVENT IV Investigators. Coronary artery bypass graft failure after on-pump and off-pump coronary artery bypass: findings from PREVENT IV. Ann Thorac Surg. 2008; 85: 494–499.