Author + information
- Received October 2, 2005
- Revision received October 28, 2005
- Accepted November 8, 2005
- Published online April 18, 2006.
- Marco Valgimigli, MD,
- Patrizia Malagutti, MD,
- Gaston A. Rodriguez-Granillo, MD,
- Héctor M. Garcia-Garcia, MD,
- Jawed Polad, MBChB, MRCP,
- Keiichi Tsuchida, MD, PhD,
- Evelyn Regar, MD, PhD,
- Willem J. Van der Giessen, MD, PhD,
- Peter de Jaegere, MD, PhD,
- Pim De Feyter, MD, PhD and
- Patrick W. Serruys, MD, PhD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Patrick W. Serruys, Thoraxcenter, Bd-406, Dr Molewaterplein 40, 3015-GD Rotterdam, the Netherlands.
Objectives This study sought to investigate whether the anatomical location of the disease carries prognostic implications in patients undergoing drug-eluting stent (DES) implantation for the left main coronary artery (LMCA) stenosis.
Background Liberal use of DES, compared with a bare metal stent (BMS), has resulted in an improved outcome in patients undergoing LMCA intervention. However, the overall event rate in this subset of patients remains high, and alternative tools to risk-stratify this population beyond conventional surgical risk status would be desirable.
Methods From April 2002 to June 2004, 130 patients received DES as part of the percutaneous intervention for LMCA stenoses in our institution. Distal LMCA disease (DLMD) was present in 94 patients. They were at higher surgical risk and presented with a greater coronary disease extent compared with patients without DLMD.
Results After a median of 587 days (range 368 to 1,179 days), the cumulative incidence of major adverse cardiac events (MACE) was significantly higher in patients with DLMD at 30% versus 11% in those without DLMD (hazard ratio [HR] 3.42, 95% confidence interval [CI] 1.34 to 9.7; p = 0.007), mainly driven by the different rate of target vessel revascularization (13% and 3%; HR 6, 95% CI 1.2 to 29; p = 0.02). After adjustment for confounders, DLMD (HR 2.79,95% CI 1.17 to 8.9; p = 0.032) and surgical risk status (HR 2.18,95% CI 1.06 to 4.5; p = 0.038) remained independent and complementary predictors of MACE.
Conclusions Distal LMCA disease carries independent prognostic implications, and it may help in selecting the most appropriate patient subset for LMCA intervention beyond the conventional surgical risk status in the DES era.
Routine implantation of drug-eluting stents (DES), by reducing the need for target vessel revascularization (TVR) and angiographic restenosis, recently has been shown to favorably affect outcome compared with bare-metal stents (BMS) in patients undergoing percutaneous left main coronary artery (LMCA) intervention (1–3). However, the rate of major cardiovascular events in the DES era remains high in the first series of patients reported (1,3). Catheter-based LMCA treatment is today mainly reserved for poor surgical candidates, which may at least partially explain the high rate of adverse events observed in this patient population. This hypothesis is based on the ability of surgical risk scores to predict both short-term and long-term outcomes in this subset of patients (1,4).
The identification of novel independent predictors of outcome beyond surgical risk status would further expand our capability to risk-stratify this patient population.
In particular, a clinical or angiographic parameter able to differentiate outcomes between percutaneous LMCA treatment and surgical revascularization would be highly desirable. This might help in selecting the appropriate subset of patients with LMCA disease in whom catheter-based treatment would be indicated, independent of surgical risk status (5).
Observational studies in the BMS era identified the distal location of the disease within the LMCA anatomy as a possible determinant of restenosis in patients undergoing percutaneous treatment of the LMCA (6). However, other investigators have not confirmed this observation (7,8), and whether this holds true in the DES era remains largely unknown.
The treatment of distal left main disease (DLMD) is offset by the need to handle the bifurcation between LMCA and the main proximal left coronary branches. The treatment of such a lesion, even with the use of DES, may remain challenging (9).
Therefore, the purpose of the present study was to investigate the clinical and angiographic outcomes of DLMD treatment in patients undergoing percutaneous revascularization in the DES era.
Study design and patient population
Since April 16, 2002, sirolimus-eluting stent (SES) implantation (Cypher, Johnson & Johnson-Cordis unit, Warren, New Jersey) have been used as a default strategy for every percutaneous coronary intervention at our institution as part of the Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) registry. From the first quarter of 2003, paclitaxel-eluting stents (Taxus, Boston Scientific, Natick, Massachusetts) became commercially available, replacing SES as the stent of choice in every percutaneous coronary intervention, as part of the Taxus-Stent Evaluated At Rotterdam Cardiology Hospital (T-SEARCH) registry. As a policy, all elective patients presenting with significant (>50% by visual estimation) LMCA disease referred to our institution for coronary revascularization are evaluated both by interventional cardiologists and by cardiac surgeons, and the decision to opt for percutaneous coronary intervention or surgery is reached by consensus, as previously described (1).
From April 16, 2002, to June 28, 2004, a total of 130 consecutive patients were treated exclusively with one or more DES in the LMCA as part of an elective or nonelective revascularization procedure and constitute the patient population of the present report. Fifty-five patients in the first cohort received exclusively SES, which were available at that time in diameters from 2.25 to 3.00 mm, whereas in the next group of 75 patients, paclitaxel-eluting stents—available in diameters from 2.25 to 3.5 mm—were implanted.
To stratify the study population into high surgical risk and low surgical risk groups, the Parsonnet surgical risk score was calculated for each patient (10). A score of >15 was used to identify patients at high risk as previously suggested (4,11). Protected LMCA segment was defined by the presence of at least one patent arterial or venous conduit to at least one left coronary segment. Nonelective treatment was defined as a procedure carried out on referral before the beginning of the next working day (12).
This protocol was approved by the hospital ethics committee and is in accordance with the Declaration of Helsinki. Written informed consent was obtained from every patient.
Procedures and post-intervention medications
All interventions were performed according to current standard guidelines. The final interventional strategy, including the use of glycoprotein IIb/IIIa inhibitors, was entirely left to the discretion of the operator, except for the stent use. Total stent length was calculated as the sum of the length of each single stent placed to treat LMCA, provided at least one stent strut was in direct contact with the left main stem at visual estimation. Angiographic success was defined as residual stenosis <30% by visual analysis and the presence of Thrombolysis In Myocardial Infarction (TIMI) flow grade 3. All patients were advised to maintain aspirin lifelong, and clopidogrel was prescribed for 6 months in both groups.
End point definitions and clinical follow-up
Distal LMCA disease was defined as significant lumen obstruction (>50%) at visual estimation occupying the third distal area of the LMCA shaft with the entire lesion or part of it either directly involving the ostium of the left anterior descending and/or circumflex artery or in close contact with at least one of them. The primary outcome was the occurrence of major adverse cardiac events, defined as: 1) death, 2) nonfatal myocardial infarction, or 3) target vessel revascularization. Patients with more than one event have been assigned the highest rank event, according to the previous list. All deaths were considered to be of cardiac origin unless a noncardiac origin was established clinically or at autopsy. Myocardial infarction was diagnosed by an increase in the creatine kinase level to more than twice the upper normal limit and with an increased creatine kinase-MB fraction. Target vessel revascularization was defined as a repeat intervention (surgical or percutaneous) to treat a luminal stenosis in the stent or within the adjacent 5-mm segments adjacent to the stent, including the ostium of the left anterior descending artery and/or circumflex artery. Information about in-hospital outcomes was obtained from an electronic clinical database for patients maintained at our institution and by review of hospital records for those discharged to referring hospitals (patients were referred from a total of 14 local hospitals). Post-discharge survival status was obtained from the Municipal Civil Registries. Data on occurrence of myocardial infarction (MI) or repeat interventions at follow-up were collected by consultation of our institutional electronic database, by contacting referring institutions, and from all living patients.
Quantitative angiographic analysis
Quantitative analyses of all angiographic data were performed with the use of edge-detection techniques (CAAS II, Pie Medical, Maastricht, the Netherlands). A value of 0 mm was assigned for the minimum luminal diameter (MLD) in cases of total occlusion at baseline or follow-up. Binary restenosis was defined as stenosis of more than 50% of the luminal diameter in the target lesion. Acute luminal gain was defined as the MLD after the index procedure minus the MLD at baseline angiography. Late loss was defined as the MLD immediately after the index procedure minus the MLD at angiographic follow-up. Net luminal gain was defined as the difference between MLD at follow-up and MLD before the procedure. Quantitative angiographic measurements of the target lesion were obtained in-stent and in-lesion (including the stented segment as well as the margins 5 mm proximal and distal to the stent).
Because the T-SEARCH is an ongoing registry at out institution, the selection of the cohort of patients for the present report was based on the following criteria: a minimum follow-up time of 1 year, and an expected major adverse cardiac events (MACE) rate of 10% in the group of patients without DLMD with a 3× event rate increase in the DLMD, based on previous findings (1), with alpha and beta errors of 5% and 20%, respectively.
Continuous variables are shown as mean ± SD and were compared using the Student unpaired ttest. Categorical variables are presented as counts and percentages and are compared with the Fisher Exact test. Survival curves were generated by the Kaplan-Meier method, and survival among groups was compared using the log-rank test. Cox proportional hazards models were used to assess risk reduction of adverse events. Multivariable analysis, considering all variables reported in Tables 1 and 2⇓with a p value of <0.10, was performed to adjust for possible confounders and to identify whether DLMD was an independent predictor of adverse events. Probability was significant at a level of <0.05. All statistical tests were two-tailed. Statistical analysis was performed on Statistica 6.1 (Statsoft Inc., Tulsa, Oklahoma).
Baseline and procedural characteristics
Baseline and procedural characteristics of the patient population, stratified into LMCA disease location, are shown in Tables 1 and 2. Patients with DLMD tended to be older and presented with an overall higher Parsonnet surgical risk score compared with those without DLMD. Similarly, coronary artery disease extent, number of stents, and total stent length were greater and the use of intravascular ultrasound was less than in DLMD patients. In one patient per group, both presenting with acute myocardial infarction, procedural success was not obtained because of TIMI flow grade <3 after stenting.
Clinical outcome based on left main disease location
At 30 days, there was no difference in clinical outcome between patients with and without DLMD, considering either the whole population, those receiving an elective intervention, or those at low surgical risk according to the Parsonnet score (Table 3).Overall, no documented thrombotic stent occlusion occurred in the first 30 days or thereafter.
After a median follow-up of 587 days (range 368 to 1,179 days; 577 days [range 368 to 1,156 days] in the DLMD group vs. 598 days [range 398 to 1,179 days] in the group without DLMD, p = 0.56), the cumulative incidence of MACE (death, MI, or TVR) was significantly higher in patients with DLMD (30% vs. 11% in those without DLMD; hazard ratio [HR] 3.42, 95% confidence interval [CI] 1.34 to 9.7; p = 0.007) (Fig. 1A).The composite death/MI was 17% in the DLMD group and 8% in patients without DLMD (HR 2.11, 95% CI 0.45 to 10; p = 0.34), whereas the cumulative incidence of TVR was 13% versus 3% in patients with and without DLMD (HR 6, 95% CI 1.2 to 29; p = 0.02) (Fig. 1B).
In the elective patient population (106 patients overall, 73 in the DLMD group), the cumulative incidence of MACE remained greater in the DLMD subgroup (26%) compared with those without DLMD (9%; HR 3.59, 95% CI 1.23 to 12.1; p = 0.01) (Fig. 1C). The composite of death/MI was 11% in patients with and 6% in those without DLMD (HR 2.48, 95% CI 0.7 to 8.5; p = 0.35), whereas the need for TVR was 15% and 3% in the two groups, respectively (HR 6.1, 95% CI 1.6 to 21; p = 0.02) (Fig. 1C). Even after excluding patients receiving protected intervention, the MACE rate remained higher in patients with compared with those without DLMD (HR 3.1, 95% CI 1.08 to 9.1; p = 0.01).
Complex bifurcation stenting was more common in the DLMD subgroup. However, the technique of stent deployment in itself failed to affect outcome, with a MACE rate of 31% in DLMD patients undergoing stenting of the main branch versus 28% in those treated with bifurcation stenting (HR 0.96, 95% CI 0.46 to 1.49; p = 0.92).
Patients at high surgical risk according to the Parsonnet score had a higher MACE rate compared with those at low risk, confirming previous findings (Fig. 2A).
To explore the additive prognostic value of combining the anatomical location of LMCA disease and surgical risk status, Kaplan-Meier curves were constructed according to the four combinations generated by having distal or nondistal LMCA disease and being at high or low surgical risk. As shown in Figure 2, the cumulative event rate in the group of patients affected by nondistal LMCA disease with low surgical risk was 10-fold lower (4%) than that observed in surgical high-risk patients with DLMD (40%, p = 0.002). Surgical high-risk status seemed to stratify patients at greater probability of early events independent of the anatomical location of LMCA disease, whereas DLMD identified patients at higher risk for late events irrespective of surgical risk.
After adjustment for Parsonnet risk score (which includes age), the extent of coronary disease, number of deployed stents, post-procedural minimal lumen diameter, bigger balloon inflated, and use of intravascular ultrasound at multivariable Cox regression analysis, DLMD remained an independent predictor of MACE (HR 2.79, 95% CI 1.17 to 8.9; p = 0.032) independent of surgical risk status (assessed as high risk versus low risk; HR 2.18, 95% CI 1.06 to 4.5; p = 0.038). The estimates of these two covariates remained unchanged if: 1) treatment technique (main branch vs. bifurcation stenting) or 2) treatment technique but not number of deployed stents—to check for possible colinearity between these two variables—were introduced in the model.
No statistical interaction emerged between the anatomical location of LMCA disease and the surgical risk with respect to MACE (p = 0.3).
Quantitative angiographic analysis
Seventy-one patients in the DLMD group (84% of eligible patients) and 28 patients without DLMD (85% of eligible patients) underwent eight-month angiographic follow-up (p > 0.99). Quantitative coronary angiography analysis is reported in Table 4.As shown, despite a similar LMCA reference vessel diameter, patients with DLMD location tended to have a longer lesion length and a slightly smaller MLD. In-stent and in-segment acute luminal gain seemed to be similar in the two groups, whereas late loss was almost double in the DLMD group. Thus, in-stent and in-segment net luminal gain was significantly lower in patients with as compared with those without DLMD.
The percutaneous treatment of LMCA disease is a challenging task, with historical event rates often reported to be unacceptably high (4,13,14). The advent of BMS has not been regarded as a major breakthrough in the percutaneous treatment of such lesions (15) because the occurrence of in-stent restenosis was believed to be associated with fatalities (4,13). Recently, the liberal use of DES to treat LMCA has been shown to favorably affect outcome compared with the use of BMS (1–3). However, the rate of major cardiovascular events in the DES era remains high in some of the first series of patients reported (1,3).
Whether the percutaneous treatment of LMCA should be strictly reserved for poor surgical candidates or offered with less restriction to patients known to be at low risk for future MACE remains highly debated (5). Ideally, risk factors able to differentiate outcomes between those undergoing catheter-based LMCA treatment and those receiving surgical revascularization might help in selecting the most appropriate revascularization strategy. Anatomical characteristics of the LMCA lesion have great potential in this regard because they may theoretically influence the percutaneous but not the surgical revascularization technique. Some early observational studies in the BMS era identified distal location of the disease with respect to LMCA anatomy as a major determinant of restenosis in patients receiving percutaneous treatment of the LMCA (6). Similarly, we and other groups have reported that the great majority of lesions, which undergo TVR at follow-up, are located in the distal tract of the LMCA (1,3,16).
The main finding of our investigation is that the long-term outcome of patients undergoing percutaneous treatment for DLMD is significantly worse compared with that of patients treated for LMCA lesions not located in the distal tract. Interestingly, this remained true in both elective and low-surgical-risk groups. The procedural success rate along with the short-term (30 days) outcome was remarkably similar between the two groups, whereas the difference between patients with and without DLMD emerged at long-term follow-up, mainly driven by a higher need for TVR in the former group. Despite the lack of statistical significance, the composite of death and nonfatal MI was consistently two times higher in the DLMD group, both in the total cohort of patients (17% vs. 8% in the non-DLMD group, p = 0.34) and after selection for elective cases only (11% vs. 6.% in the non-DLMD group, p = 0.35). Whether these results on death and MI reflect a type II error or a chance finding remains unclear.
Our multivariable model, based on all possible confounders of clinical outcome, showed that DLMD is an independent predictor of poor outcome in this subset of patients, with an adjusted risk for MACE of almost three-fold higher than that for non-DLMD at long-term follow-up.
To further rule out the possibility that treatment technique, more than the anatomical location of the disease, was responsible for the difference in long-term outcomes between patients with and without DLMD, the MACE rate in patients receiving bifurcation stenting was compared with that in patients undergoing single-vessel stenting in the DLMD subgroup. The MACE rate was remarkably similar between these two groups of patients, supporting the hypothesis that the anatomical location of the LMCA disease more than the techniques used to treat it was responsible for the higher overall event rate in the DLMD group.
In the quantitative angiographic analysis, the late loss was higher, the acute luminal gain was lower, and the binary restenosis rate also tended to be higher in the DLMD group, which provides a possible mechanistic explanation for our clinical findings.
Finally, to investigate the prognostic power of DLMD in relation to surgical risk status, Kaplan-Meier curves were constructed according to the four combinations generated by having distal or nondistal LMCA disease and being at high or low surgical risk. Surgical risk status seemed to stratify patients at a greater probability of events independent of the anatomical location of LMCA disease, whereas DLMD identified patients at a higher risk for poor prognosis irrespective of surgical risk. Of note is that the surgical risk status seemed to better risk-stratify patients according to early events, with survival curves running parallel after the first month of treatment. Conversely, DLMD identified patients with a higher event rate at follow-up, with the two survival curves diverging at around 180 days after the procedure. Interestingly, we failed to identify a statistical interaction between surgical risk status and the location of LMCA disease in our patient population, which implies that the risk associated with DLMD is additive—not multiplicative—with respect to that of being at high surgical risk status. This further confirms that the two currently used approaches for risk stratification, namely surgical risk status and anatomical location of the disease, are independent and possibly complementary to each other.
As a potential corollary to our findings, the extremely low event rate in patients at low surgical risk undergoing LMCA intervention for nondistal LMCA disease should not go unnoticed. This subset of patients with excellent long-term outcome after catheter-based treatment should be ideally selected to prospectively test whether percutaneous intervention is a valuable alternative to surgical revascularization in future trials.
A limitation of proposing distal location of LMCA disease as a risk-stratifying tool lies in the recognition that the distal site is the most prevalent site of disease in LMCA patients undergoing catheter-based intervention in the DES era. Whether this partially reflects the fact that patients with DLMD are more likely to be at higher surgical risk or tend to be older, as in our present series, and consequently are more likely to undergo percutaneous instead of surgical revascularization, remains to be addressed.
The results of our study are encouraging, but they cannot be conclusive. Studies with larger sample sizes and more prolonged clinical follow-up are clearly in demand to confirm our findings and extend our capability to risk-stratify this challenging subset of patients.
The percutaneous treatment of DLMD emerged as a major predictor of poor long-term outcome, independent of the type of procedure (elective versus nonelective) and the overall surgical risk status. Conversely, the event rate after treatment for non-DLMD seemed to be remarkably low, with an excellent short-term and long-term prognosis, especially in the low surgical risk population. Our current findings extend the previous knowledge about risk stratification for patients undergoing catheter-based treatment of LMCA, and may help in identifying the most appropriate LMCA population, in which catheter-based intervention may be indicated beyond surgical risk status in the DES era.
- Abbreviations and Acronyms
- bare-metal stent
- drug-eluting stent
- distal left main disease
- left main coronary artery
- major adverse cardiac events
- myocardial infarction
- minimal luminal diameter
- sirolimus-eluting stent
- Thrombolysis In Myocardial Infarction
- target vessel revascularization
- Received October 2, 2005.
- Revision received October 28, 2005.
- Accepted November 8, 2005.
- American College of Cardiology Foundation
- Valgimigli M.,
- van Mieghem C.A.,
- Ong A.T.,
- et al.
- Park S.J.,
- Kim Y.H.,
- Lee B.K.,
- et al.
- Chieffo A.,
- Stankovic G.,
- Bonizzoni E.,
- et al.
- Takagi T.,
- Stankovic G.,
- Finci L.,
- et al.
- Baim D.S.
- Colombo A.,
- Moses J.W.,
- Morice M.C.,
- et al.
- Roques F.,
- Nashef S.A.,
- Michel P.,
- et al.
- Ellis S.G.,
- Tamai H.,
- Nobuyoshi M.,
- et al.
- Tan W.A.,
- Tamai H.,
- Park S.J.,
- et al.
- Agostoni P.,
- Valgimigli M.,
- van Mieghem C.A.,
- et al.