Author + information
- Received February 12, 1998
- Revision received June 22, 1998
- Accepted August 6, 1998
- Published online December 1, 1998.
- Harold L. Dauerman, MDa,
- Peter J. Higgins, MDa,
- Anthony M. Sparano, BS∗,
- C.Michael Gibson, MD, MS∗,
- Gary R. Garber, MD, FACCa,1,
- Joseph P. Carrozza Jr., MD, FACCa,
- Richard E. Kuntz, MD, MSca,
- Roger J. Laham, MDa,
- Samuel J. Shubrooks Jr., MD, FACCa,
- Donald S. Baim, MD, FACCa and
- David J. Cohen, MD, MSca,* ()
- ↵*Address for correspondence: Dr. David J. Cohen, Cardiovascular Division, Beth Israel–Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts 02215
Objectives. The purpose of this study was to compare the immediate angiographic and long-term results of debulking versus balloon angioplasty for treatment of true bifurcation lesions.
Background. Previous studies have shown true bifurcation lesions to be a high risk morphological subset for percutaneous transluminal coronary angioplasty (PTCA). Although atherectomy devices have been used to treat bifurcation lesions, no studies have compared the outcomes of these alternative treatment modalities.
Methods. Between January 1992 and May 1997, we treated 70 consecutive patients with true bifurcationlesions (defined as a greater than 50% stenosis in both the parent vessel and contiguous side branch) with conventional PTCA (n = 30) or debulking (with rotational or directional atherectomy) plus adjunctive PTCA (n = 40). Paired angiograms were analyzed by quantitative angiography, and clinical follow-up was obtained in all patients.
Results. Acute procedural success was 73% in the PTCA group and 97% in the debulking group (p = 0.01). Major in-hospital complications occurred in two patients in the PTCA group and one in the debulking group. Treatment with atherectomy plus PTCA resulted in lower postprocedure residual stenoses than PTCA alone (16 ± 15% vs. 33 ± 17% in the parent vessel, and 6 ± 15% vs. 39 ± 22% in the side branch; p < 0.001 for both comparisons). At 1 year follow-up, the incidence of target vessel revascularization (TVR) was 53% in the PTCA group as compared with 28% in the debulking group (p = 0.05). Independent predictors of the need for repeat TVR were side branch diameter >2.3 mm, longer lesion lengths, and treatment with PTCA alone.
Conclusions. For the treatment of true bifurcation lesions, atherectomy with adjunctive PTCA is safe, improves acute angiographic results, and decreases target vessel revascularization compared to PTCA alone. The benefits of debulking for bifurcation lesions were especially seen in lesions involving large side branches.
True bifurcation lesions occur in 2% to 16% of stenoses being considered for percutaneous coronary intervention (1–3). Such bifurcation morphology has been shown consistently to be associated with reduced acute success and higher complication rates after treatment by conventional balloon angioplasty (2,4,5). The higher rate of complications may relate to plaque shifting, ostial recoil and propagation of dissection (1,5)as well as endothelial dysfunction occurring specifically at coronary bifurcations (6). For this reason, multiple approaches have been used to treat such lesions including double wire and kissing balloon angioplasty (3,7–9), directional atherectomy (1,10–13), rotational atherectomy (14), and most recently complex applications of coronary stents (15–19). To date, however, there have been no direct comparisons of balloon angioplasty and other treatment modalities for patients with such lesions in order to establish the relative merits of newer devices compared to balloon angioplasty alone. For this reason, we analyzed in-hospital results and long-term outcomes for 70 consecutive patients with true bifurcation lesions treated with either conventional PTCA (percutaneous transluminal coronary angioplasty) or debulking (with rotational or directional atherectomy) plus adjunctive PTCA.
Between January 1, 1992, and May 31, 1997, a total of 70 patients underwent percutaneous coronary revascularization at Beth Israel–Deaconess Medical Center for management of symptomatic true bifurcation lesions. This represented 1.2% of all coronary interventions during that time period. Only lesions in which there was a >50% diameter stenosis in both the parent vessel and the ostium of the contiguous side branch with reference diameter >2.0 mm were considered. Of these 70 patients, 30 were treated with balloon angioplasty alone (PTCA group), and 40 were treated with a combination of either rotational atherectomy or directional atherectomy (DCA) and adjunctive PTCA (debulking group). In general, PTCA was performed in the earlier portion of the study period (1992 to 1994) and debulking was performed during the later years (1995 to 1997).
As previously described, PTCA for true bifurcation lesions was performed using a double-wire technique in which simultaneous (“kissing”) or sequential balloon inflations were performed in both vessels (7–9,20). Directional atherectomy was our treatment of choice in larger noncalcified vessels and was performed according to standard techniques without the use of a nitinol wire for side branch protection (10–12,21). A series of four to eight circumferential cuts were performed at an inflation pressure of 10 to 40 psi, first in the parent and then in the side branch. If the vessel was greater than 3.2 mm in diameter, we generally used a 7F atherectomy catheter. If the vessel was 2.5 to 3.2 mm in diameter, a 6F atherectomy catheter was used. Rotational atherectomy was performed sequentially in the parent and side branch using an alternating “stepped burr” approach, with placement of the wire sequentially in both branches for each sized burr (14,22). We generally began with a 1.5- to 1.75-mm burr and increased in 0.25-mm to 0.5-mm increments to a final burr size corresponding to 60% to 80% of each reference vessel diameter.
Following completion of either rotational or DCA, adjunctive kissing balloon angioplasty was performed in all cases, generally at low inflation pressures (2 to 8 atm). Vascular access sheaths were removed once the activated clotting time was ≤180 s. Intravenous heparin was then restarted and administered overnight. Myocardial infarction was diagnosed if any postprocedure creatine kinase measurement was ≥2 times the upper limit of normal (200 IU/ml) with a positive MB fraction or if new Q waves appeared on the 12-lead electrocardiogram.
Angiographic analysis and clinical follow-up
Quantitative analysis of the coronary segments was performed using a validated, automated edge-detection algorithm (23), with the dye-filled catheter as a reference. The diameter of the normal segment proximal to the traced area in the parent vessel was used to determine the parent reference diameter, and the side branch reference diameter was determined from the diameter of the traced area in the normal segment distal to the lesion in the branch. The minimal luminal diameter, reference diameter and percent stenosis were calculated as the mean values from multiple projections. Lesion length was defined as the distance from the proximal to the distal shoulder of the lesion. Thrombolysis in Myocardial Infarction trial (TIMI) grade flow was assessed by frame counts using a digital frame counter, as previously described (24).
Bifurcation lesion type was classified according to the Duke system (25). Class D bifurcations were defined as >50% diameter stenosis in the proximal and distal parent vessel and a >50% stenosis in the ostium of the contiguous side branch (Fig. 1A). Class F bifurcations were defined as >50% diameter stenosis in the proximal parent vessel, and >50% diameter stenosis in the contiguous side branch ostium but no significant stenosis of the parent vessel beyond the bifurcation (Fig. 1B). Side branch compromise was defined as abrupt closure or decrease in TIMI grade flow in the side branch at any time during the procedure. Procedural success was defined as achievement of a residual stenosis <50% in both the parent vessel and side branch in the absence of any major complications (death, Q wave myocardial infarction [MI] or repeat revascularization) prior to hospital discharge.
Clinical follow-up was obtained at 1 month, 6 months, and 1 year after treatment and annually thereafter to determine each patient’s vital status and need for subsequent revascularization. Target vessel revascularization (TVR) was defined as any repeat percutaneous intervention to the target vessel (parent or side branch) or any coronary bypass graft to the treated vessel during follow-up.
Discrete data are presented as frequencies, and continuous data as mean ± SD. Continuous data were compared using Student’s ttest, and frequencies were compared using the Fisher exact test. Estimates of event-free survival and freedom from TVR were determined using the Kaplan-Meier method (26)and are displayed as standard survival curves. The univariate relationship between individual covariates and the need for subsequent target vessel revascularization was assessed using the log-rank statistic. A p value ≤0.05 was considered statistically significant. A multivariable proportional hazards model (27)was used to identify the independent effect of treatment strategy on TVR while adjusting for baseline differences in vessel size, lesion length, diabetes and additional preprocedure variables identified as predictors of TVR in the univariate analysis. Finally, stepwise regression analysis was performed to identify those variables associated with repeat TVR. Factors considered for the stepwise regression model included patient age, gender, hypercholesterolemia, diabetes mellitus, smoking status, vessel treated, reference vessel diameter (parent and branch), lesion length (parent and branch), minimum lumen diameter (MLD) (parent and branch), bifurcation type, calcification, use of adjunctive abciximab and treatment strategy. All analyses were performed using the STATA 3.0 statistical package (Computing Resource Center, Santa Monica, California).
The treatment groups were generally well matched with respect to age, gender, risk factors and lesion location (Table 1). Preprocedure angiographic characteristics were also well matched between the two treatment groups (Table 2). Approximately 70% of the patients had class D bifurcation lesions (“Mercedes-Benz” configuration) and 30% had class F lesions according to the Duke classification scheme (Fig. 1)(25). There were no differences in the baseline lesion severity or flow velocity as assessed by the TIMI frame count. There were no differences in operator experience in the debulking (7.4 ± 5.3 years) versus PTCA (7.6 ± 3.9 years) groups (p = 0.86).
In the debulking group, 34 (85%) patients underwent DCA and 6 patients received rotational atherectomy (15%). Rotational atherectomy was performed with 1 to 2 burrs with a size range of 1.5 to 2.15 mm in the parent branch, and 1.25 to 1.75 mm in the side branch. Directional coronary atherectomy was performed with a 7F device in 27 parent vessels (79%) and in 23 side branches (68%); the remainder of the lesions were treated with a 6F device. Two patients in the debulking group received stents in the parent vessel, one owing to inadequate lesion expansion, and the other to a moderate parent vessel dissection. The final balloon size was similar in the debulking and PTCA groups in both the parent vessel (3.2 ± 0.7 mm vs. 3.0 ± 0.4 mm, p = 0.10) and side branch (2.8 ± 0.9 vs. 2.5 ± 0.4, p = 0.10). The maximum inflation pressures were significantly lower in the debulking group for both the parent vessel (5.1 ± 3.3 vs. 7.0 ± 1.8 atm, p = 0.007) and side branch (4.0 ± 2.4 vs. 7.0 ± 2.3 atm, p = 0.001). Abciximab was administered to seven patients in the debulking group compared to none in the PTCA group (p = 0.04).
In-hospital results and complications
Procedural success was 73% in the PTCA group and 97% in the debulking group (p = 0.01). This difference was mainly due to failure to achieve a <50% residual stenosis in the side branch of seven (23%) patients in the PTCA group. There were no deaths in either group. In the debulking group, one patient had a coronary dissection that led to abrupt closure, MI and emergent coronary artery bypass grafting (CABG). In the PTCA group, one patient had a Q wave MI on the day following the procedure, and another patient had abrupt closure 2 days after the procedure leading to emergent CABG.
In the debulking group, transient side branch compromise occurred in eight patients (22%) as compared with only one patient (3%) in the PTCA group (p = 0.06). Each episode of side branch compromise in the debulking group was transient, as the snow-plowed branches could invariably be rescued (despite the absence of a protecting guide wire). In fact, all patients in the debulking group had TIMI-III flow at the completion of the procedure. Moreover, despite more frequent transient side branch loss in the debulking group, there was no difference in periprocedural MI between the two groups (5% vs. 3%, p = NS).
Acute angiographic results (Table 3)
The MLD of the parent vessel increased from 0.61 ± 0.30 mm to 2.09 ± 0.65 mm after initial debulking alone, and further increased to 2.48 ± 0.57 mm after adjunctive PTCA (Fig. 2, p < 0.001 for both comparisons). Similar benefits were seen for the branch vessel as well. The final MLDs in the parent vessel (2.48 ± 0.57 vs. 1.97 ± 0.65 mm) and the branch vessel (2.22 ± 0.53 vs. 1.39 ± 0.55 mm) were substantially larger in the group treated with initial debulking as compared with PTCA alone (p < 0.001 for both comparisons). Patients treated with initial debulking also had less frequent persistent dissections in the parent vessel or side branch (9% vs. 33%, p = 0.03) and improved coronary flow velocities in the side branches as judged by the TIMI frame counts (19 ± 5 vs. 24 ± 12, p = 0.03).
Late clinical results
Clinical follow-up was available for all patients with a median duration of 2.0 years in the debulking group and 3.8 years in the PTCA group with 6-month follow-up available for 68 of 70 (97%) of patients. According to Kaplan-Meier analysis, there was no significant difference in mortality or the occurrence of MI between the PTCA and debulking groups, respectively (Table 4). In contrast, the need for subsequent TVR was significantly reduced in patients treated with debulking plus adjunctive PTCA as compared with balloon angioplasty alone (Fig. 3). At 1-year follow-up, the incidence of repeat TVR was 28% in the debulking group as compared with 53% for the PTCA group (p = 0.05). The extent of TVR differed between the two treatment groups as well. While most patients in our series who underwent TVR required treatment of at least the parent vessel, patients in the PTCA-only group were much more likely to require retreatment of the sidebranch as well (100% vs. 54%, p = 0.02).
Stepwise multivariable regression analysis identified greater lesion length and sidebranch diameter >2.3 mm as preprocedure characteristics associated with late TVR (Table 5). After adjusting for baseline differences in these characteristics, treatment with debulking remained an independent predictor of late TVR with a relative risk of 0.4 (95% CI, 0.2 to 1.0). Neither the use of adjunctive abciximab nor the type of debulking procedure (DCA vs. rotational atherectomy) had an influence on late TVR. The benefits of debulking were most pronounced for patients with large side branches (Fig. 4). For those patients with side branch diameters >2.3 mm, the 6-month incidence of TVR was 29% with debulking and 64% with simple PTCA (p = 0.03).
Significant improvements in device technology over the past decade have allowed the expanded use of percutaneous coronary interventions in higher risk clinical scenarios and specific morphological subsets that previously were associated with poor outcomes with balloon angioplasty alone (28). One such subset is bifurcation coronary lesions, which have been shown consistently to be associated with higher rates of acute complications and clinical restenosis when treated with PTCA (2,4,5). Consequently, a variety of new devices including DCA, rotational atherectomy, and various coronary stent configurations have been used in an attempt to treat these challenging lesions (10–19). To date, however, there have been no studies comparing the acute or long-term results of these techniques.
In this study, we compared the in-hospital and long-term outcome of treatment of true bifurcation lesions with PTCA or a debulking strategy using atherectomy followed by adjunctive PTCA. We found that treatment of true bifurcation lesions with mechanical debulking techniques was safe compared with conventional balloon dilation. Moreover, debulking followed by adjunctive postdilation was associated with a significant improvement in procedural success (97% vs. 73%, p = 0.01) and the attainment of larger postprocedure lumen diameters in both the parent and branch vessel. Most importantly, we found that the debulking strategies resulted in improved late clinical outcomes, with a reduction in the need for repeat TVR at 1-year follow-up from 53% to 28% (p = 0.05).
Comparison with previous studies
Our results are generally comparable to the results of previous small series of the broader population of bifurcation lesions. Studies of PTCA for bifurcation lesions have reported acute procedural success rates of 74% to 94% (2,5,9)with residual stenoses of 30% to 40% in both the parent vessel and side branch (5,20,29)—similar to the results we observed. Several previous studies have reported clinical restenosis rates of 37% to 42% after PTCA for bifurcation lesions (3,9). The fact that our PTCA results are somewhat inferior to these historical results may relate to the inclusion of only the highest risk subset of bifurcation lesions in our series. Most prior studies of “bifurcation” lesions have included many patients without a significant stenosis in a major contiguous side branch (3,4,20,29–33)—lesions with a lower risk of acute and long-term complications than true bifurcations (30,31,33).
Previous studies of DCA or rotational atherectomy have demonstrated that these techniques can be accomplished safely in selected bifurcation lesions (11–14). In the largest series reported to date, Lewis and colleagues used DCA with a double-wire technique to achieve procedural success in 97% of patients, with a residual stenosis of 12% in the parent vessel and 17% in the side branch (11)—similar to the results we obtained. In the original multicenter Rotablator experience, bifurcation lesions comprised 27% of the registry, and procedural success rate for these lesions was 95%; how many of these were true bifurcation lesions treated with ablation in both the parent vessel and side branch is unclear (14,22).
Our study serves to expand our understanding of the percutaneous treatment of bifurcation coronary lesions in several ways. First, this is the largest series to date of patients treated with atherectomy techniques for true bifurcation lesions. Moreover, this is the first study to directly compare the acute results of treatment for bifurcation lesions with both conventional PTCA and atherectomy techniques. Finally, by examining long-term outcomes, ours is the first study to clearly demonstrate the clinical benefits of debulking techniques over conventional PTCA for such patients.
Safety of atherectomy without side branch protection
Previous studies have suggested that debulking—primarily with DCA—is associated with a higher incidence of side branch loss and associated periprocedural infarction than PTCA (29). These observations have led to the development of bifurcation atherectomy techniques using a second nitinol guide wire to protect the adjacent side branch (11,13). Our experience with true bifurcation lesions suggests that, with careful attention to technique, treatment with debulking techniques can be performed safely without the need to protect the side branch. Consistent with previous studies, we observed a higher incidence of transient side branch compromise with debulking than with conventional PTCA (22% vs. 3%). Nonetheless, in all cases the side branch was successfully rescued, and in no case did transient side branch loss lead to periprocedural infarction. These findings are in contrast to the experience in the CAVEAT trial, in which side branch compromise occurred in 15% of atherectomy patients and was associated with a 9% incidence of periprocedural infarction (29). It is likely that many of the side branches occluded in CAVEAT were relatively small and not aggressively rescued, however, thus contributing to the relatively high incidence of periprocedural infarction.
Benefits of debulking
It should not be surprising that debulking resulted in substantially better acute angiographic results than conventional PTCA in true bifurcation lesions. In the BOAT trial, DCA resulted in a mean residual stenosis of 15% as compared with 28% for PTCA (34). These results are virtually identical to those we observed in the parent vessel in this study. The benefits of debulking were even more striking for the side branches, where the average acute gain was more than twice that achieved with PTCA alone (1.44 vs. 0.69 mm, p < 0.001). These benefits likely reflect the ability of atherectomy techniques to limit both elastic recoil and “plaque shifting,” both of which frequently compromise the results of simple balloon dilation for ostial disease. Given the significant differences in acute angiographic results between the two treatment groups, the improved clinical outcome seen with debulking in this study should not be surprising. Previous studies have identified postprocedure MLD as the principal determinant of restenosis after PTCA (35), DCA (34–36), excimer laser angioplasty (37), and coronary stenting (38). In the case of bifurcation lesions, the ability of two separate lesions to contribute independently to the process of clinical restenosis would be expected to magnify this relationship.
In addition to reducing the need for repeat TVR in patients with bifurcation lesions, the use of debulking techniques appears to reduce the complexity of subsequent interventions as well. Among patients who developed clinical restenosis in our study, patients in the debulking group were less likely to require treatment of both the parent and side branch at follow-up as compared with those patients in the PTCA group (54% vs. 100%, p = 0.02). This led to simple percutaneous management of restenosis (frequently with stenting of the parent vessel only) in 10 of 13 patients in the debulking group who required repeat TVR. In contrast, those patients treated with initial PTCA who returned with clinical restenosis invariably required treatment of both the parent vessel and side branch, leading to more complex second interventions with higher rates of recurrent restenosis and subsequent bypass surgery. As a result, the 1-year incidence of CABG was significantly lower for patients treated with initial debulking as opposed to PTCA (2% vs. 21%, p = 0.05).
The major limitation of our study is its retrospective, observational design with possible confounding by baseline differences in patient characteristics. The fact that debulking was retained in our multivariable analysis as an independent predictor of freedom from TVR, however, lends considerable support to the validity of this finding. Our study is further limited by the lack of routine angiographic follow-up. Nonetheless, the ultimate goal of percutaneous coronary revascularization is relief of angina and avoidance of additional procedures, both of which were accomplished more frequently with the debulking strategy. Finally, our study is unable to address questions as to the potential benefits of coronary stenting for bifurcation stenoses (15–19). However, the application of current stent designs to bifurcation lesions poses a number of technical challenges that frequently lead to unstented gaps, difficulties with simultaneous stent delivery, and possibly increased risks inherent in the use of multiple stents (6,39). Given the results of atherectomy for such patients, future studies of the outcomes of stenting for bifurcation lesions should be compared with the excellent results achieved using debulking techniques.
In this observational study, we found that for treatment of true bifurcation lesions, a strategy of initial debulking with either DCA or rotational atherectomy followed by adjunctive PTCA resulted in improved procedural success and a lower need for repeat TVR during follow-up, especially in patients with larger side branches. These findings suggest that optimal treatment for bifurcation lesions should be defined based on the size of the involved sidebranch. For patients with a side branch >2.3 mm in diameter, initial debulking with directional or rotational atherectomy followed by adjunctive balloon dilation appears to produce the best long-term outcome. If the side branch is smaller, however, the benefits of debulking on late outcome appear minimal. In this case, primary definitive treatment of the parent vessel with PTCA or stenting would appear reasonable, with subsequent rescue of the side branch if necessary.
↵1 Dr. Garber is deceased.
☆ Dr. Cohen was supported in part by a Clinician-Scientist Award from the American Heart Association, Dallas, Texas.
- coronary artery bypass grafting
- directional atherectomy
- myocardial infarction
- minimum lumen diameter
- percutaneous transluminal coronary angioplasty
- target vessel revascularization
- Received February 12, 1998.
- Revision received June 22, 1998.
- Accepted August 6, 1998.
- American College of Cardiology
- Myler R.K.,
- Shaw R.E.,
- Stertzer S.H.,
- et al.
- The Multivessel Angioplasty Prognosis Study Group,
- Ellis S.G.,
- Vandormael M.G.,
- Cowley M.J.,
- et al.
- McLenachan J.M.,
- Vita J.,
- Fish R.D.,
- et al.
- Mansour M.,
- Fishman R.F.,
- Kuntz R.E.,
- et al.
- Reisman M.
- Kobayashi Y.,
- Colombo A.,
- Reimers B.,
- Di Mario C.
- Oda H.,
- Ito E.,
- Miida T.,
- Toeda T.,
- Higuma N.
- Safian R.D.,
- Gelbfish J.S.,
- Erny R.E.,
- Schnitt S.J.,
- Schmidt D.A.,
- Baim D.S.
- Warth D.C.,
- Leon M.B.,
- O’Neill W.,
- Zacca N.,
- Polissar N.L.,
- Buchbinder M.
- Gibson C.M.,
- Cannon C.P.,
- Daley W.L.,
- et al.
- Popma J.J.,
- Leon M.B.,
- Topol E.J.
- Cox D.R.
- Baim D.S.,
- Cutlip D.E.,
- Sharma S.K.,
- et al.
- Kuntz R.E.,
- Gibson C.M.,
- Noboyushi M.,
- Baim D.S.
- Bittl J.A.,
- Kuntz R.E.,
- Estella P.,
- Sanborn T.A.,
- Baim D.S.