This retrospective observational study collected informations from all patients who underwent percutaneous treatment of a de novo coronary bifurcation lesion in a single German center between January 2008 and August 2011. We included all consecutively presenting patients with a bifurcation lesion where the side branch had been covered with a stent placed inside the main branch.
The bifurcation lesions were defined according to the Medina classification .
The study population included male or female patients older than 18 years of age with a diagnosis of stable angina or silent ischemia. Patients with an acute coronary syndrome (unstable angina, NSTEMI, STEMI, cardiogenic shock) were not excluded. Patients with a primary occlusion of the target vessel and a significant bifurcation lesion visible after thrombectomy were also kept in the patient collective.
The following patients were excluded for methodological reasons: patients with an in-stent-restenosis, patients with a therapy using a drug-coated balloon during the procedure (inside the main branch and/or the side branch) and patients where the side branch had not been covered by the stent inside the main branch.
Two different treatment strategies were compared: group A represented patients with a simple strategy without any treatment of the side branch (balloon angioplasty or stenting). Group B consisted of patients where the operator treated the side branch (balloon angioplasty and/or stenting) after or before stenting of the main branch following the concept of “provisional stenting”.
The demographic data, patient history, coronary risk factors, lesion location, morphology and procedural strategy were all documented.
For all patients we used a systematic approach for treating patients with coronary bifurcation lesions. This standard was established before the initiation of this registry. This standard is described in the chapter “angiographic procedure”.
The study complied with the Declaration of Helsinki regarding research on humans. All patients provided their written informed consent. An approval of an ethics committee was not intended due to the retrospective nature of our study.
Patients with an acute coronary syndrome were treated with aspirin 500 mg intravenously and 5000 IE heparin before admission to our hospital. If the procedure was elective patients were preloaded with 300 mg clopidogrel.
After the procedure, patients were maintained on aspirin 100 mg and clopidogrel 75 mg daily. After BMS implantation clopidogrel was used for 4 weeks, after DES implantation it was used for 6 months, and in case of an acute coronary syndrome it was used for 12 months after the index procedure. Life-long aspirin was prescribed for every patient. Other medications such as ß-Blockers, angiotensin-converting enzyme inhibitors and statins were given as indicated.
Use of glycoprotein IIb/IIIa inihibitors was left to the discretion of the operator. GPIIb/IIIa inhibitors represents a component of standard care at our hospital wherever a large thrombus burden is observed.
During the procedure, intravenous heparin was given to maintain an activated clotting time of > 250 sec. The radial or femoral approach and 6 F guiding catheters were used as a matter of routine.
Our standard procedure (established before the start of this registry) during interventions for bifurcation lesions involved the following: (1) wiring of both branches of the bifurcation lesion, (2) pre-dilatation of the main branch, (3) stenting of the main branch using a DES or a BMS, (4) in case of residual stenosis of the side branch > 50% or a TIMI flow < 2 of the side branch angioplasty or stenting (“provisional stenting”), (5) final kissing wherever a side branch had an angioplasty or stenting. Patients with a diameter of the side branch < 1.5 mm by eye ball were not included.
If a good result was achieved with a TIMI flow III in both branches after stenting of only the main branch, or if there was no visible stenosis of the side branch > 50% and/or the patient had no angina after removal of the balloon from the lesion the procedure was terminated without PCI of the side branch. In this group a really simple concept was scheduled, with no attempted treatment strategy for the side branch. These patients are represented in group A.
A defined treatment strategy for the side branch was prespecified only in patients with the above mentioned criterias (group B, e.g. TIMI flow reduction or a visible stenosis of the side branch > 50%).
A final kiss was attempted for every patient where the side branch was treated, but this was not the case where patients were treated using the simple strategy (group A). The first-line therapy in this case was a simultaneous balloon angioplasty as a “final kissing” PCI. If this strategy could not be accomplished due to failure of balloon placement, the strategy was switched to a sequential balloon angioplasty.
Indications for the placement of a side branch stent in group B included  a residual stenosis after balloon angioplasty > 50%,  a flow limiting dissection,  presence of a thrombus or  occlusion of the side branch after balloon angioplasty.
Implantation of additional stents to cover the whole lesion or a dissection were also permitted. However, no combination of drug-eluting stents with bare metal stents was to have been performed.
The balloons were choosen to achieve a balloon-vessel ratio of ~ 1 after measuring the pre-procedure reference vessel diameter. We used semi-compliant balloons with nominal inflation pressures.
Implantation of a drug-eluting stent was preferred in patients with an acute coronary syndrome (NSTEMI and STEMI), in patients with long lesions (> 20 mm) and a small vessel diameter (< 2.5 mm), and in diabetes mellitus patients.
Quantitative coronary angiographic (QCA) analysis was performed using Quantcor QCA (version V2.0 by Pie Medical Imaging, Maastricht, The Netherlands). QCA measurements were performed by an independent operator unaware of the details of the therapy using a dye-filled catheter as reference.
The minimal lumen diameter (MLD) and percentage diameter stenosis were measured pre-procedure and post-procedure. The reference vessel diameter of the main branch and the side branch was set up where the diseased segment seems to be unobtrusive.
Clinical definitions and follow-up
Follow-up was carried out either at an office visit, by looking for data in our local hospital data base or via a telephone call (with the patient or the home physician) where the rate of MACE (major adverse cardiac events) was determined. No scheduled follow-up angiography was indicated unless it was for patients with significant coronary lesions for PCI not related to the target bifurcation lesion, or patients with evidence of disease progression due to a new angina pectoris and/or objective evidence of ischemia.
The follow-up data were defined as either death, myocardial infarction (STEMI and NSTEMI), stent thrombosis, CABG or target lesion revascularization (TLR). All deaths were considered cardiac unless otherwise documented.
The diagnosis of AMI (STEMI or NSTEMI) both peri-procedurally and at follow-up required an elevation of creatine kinase to levels twice those of the upper normal limit together with a rise in the creatine kinase-MB fraction, an elevation of troponin I and/or new ST-segment elevations or new Q-waves (ECG). The threshold used for classifying a positive troponin I test was 0.1 ng/ml. For CK the manufacturer reported a lower threshold of > 2.8 μmol/s/l.
Target lesion revascularization (TLR) was defined as either surgical or percutaneous re-intervention driven by 1) significant (> 50%) luminal diameter narrowing either within the stent or within the 5 mm proximal or distal to the MB or SB stent edge, 2) stent thrombosis or 3) TLR-related CABG. This was undertaken in the presence of either anginal symptoms or objective evidence of an ischemia including stent thrombosis.
Target vessel revascularization (TVR) was defined as revascularization by PCI or surgery within the target vessel encompassing the target lesion including TLR plus PCI target vessel-non target lesion plus CABG including target vessel, not related to the target lesion.
MACE was defined as TVR plus cardiac death.
Classification of a stent thrombosis was based on the definitions of the Academic Research Consortium (ARC) regarding definite, probable or possible stent thrombosis . Stent thromboses were categorized according to the timing of the event into: intraprocedural thrombosis, subacute thrombosis (from the end of the procedure to 30 days), and late stent thrombosis (> 30 days).
95% confidence intervals for the differences between both groups are presented in the tables. For binary variables, odds ratios and their 95% corresponding confidence intervals were calculated using logistic regressions; for ordinate variables, confidence intervals referred to the differences between the means.
For a few variables, as indicated in the tables, log values were taken to accommodate for deviations from normality where the differences in the resulting means were tested using t-tests. Statistics from all tests are reported as 2-sided probability values. All calculations were performed using Stata 11.