Hybrid Coronary Revascularization in the Era of Drug-Eluting Stents Gavin J

REVIEW

Hybrid Coronary Revascularization in the Era of Drug-Eluting Stents Gavin J. Murphy, FRCS, Alan J. Bryan, FRCS, and Gianni D. Angelini, FRCS Bristol Heart Institute, University of Bristol, Bristol, United Kingdom

Left internal mammary artery to left anterior descending the costs and logistic problems associated with perform- coronary artery bypass grafting integrated with percuta- ing two procedures with differing periprocedural man- neous coronary angioplasty (hybrid procedure) offers agement protocols. Recently, drug-eluting stents have multivessel revascularization with minimal morbidity in reduced the need for repeat intervention after percutane- high-risk patients. This is caused in part by the avoid- ous coronary intervention, and this has raised the possi- ance of cardiopulmonary bypass–related morbidity and bility that the results of hybrid revascularization may manipulation of the aorta coupled with minimally inva- now equal or even better those of off-pump coronary sive techniques. Hybrid revascularization is currently artery bypass grafting. Although undoubtedly effective reserved for particularly high-risk patients or those with at reducing in-stent restenosis, drug-eluting stents will favorable anatomic variants however, largely because of not address the issues of incomplete revascularization or the emergence of off-pump coronary artery bypass graft- the logistic problems associated with hybrid. Uncertainty ing, which permits more complete multivessel revascu- regarding the long-term effectiveness of drug-eluting larization, with low morbidity in high-risk groups. The stents in many patients, as well as their high cost when wider introduction of hybrid revascularization is limited compared with those of off-pump coronary artery bypass chieﬂy by the high number of repeat interventions com- grafting surgery, also militates against the wider intro- pared with off-pump coronary artery bypass grafting, duction of hybrid revascularization. which occurs because of the target vessel failure rate of (Ann Thorac Surg 2004;78:1861–7) percutaneous coronary intervention. Other demerits are © 2004 by The Society of Thoracic Surgeons

ybrid or integrated coronary revascularization strat- supersede hybrid revascularization was based on the fact H egies evolved in the mid 1990s as a result of the that multivessel CABG had lower reintervention rates desire to effectively treat patients with multivessel dis- than multivessel PCI [2]. Off-pump CABG was also REVIEW ease while at the same time lowering procedure-related shown to result in superior short-term and equivalent morbidity by combining minimal access coronary sur- midterm outcomes compared with conventional CABG, gery with percutaneous techniques [1]. At the time this at a reduced cost [3–6]. More recently, the apparent was heralded as innovative, representing as it did a success of drug-eluting stents (DES) at reducing reinter- blurring of the boundaries between interventional cardi- vention rates compared with PCI using non-DES, cou- ology and cardiac surgery. Hybrid revascularization rep- pled with the desire of interventional cardiologists to resented the natural evolution of the two disciplines, in broaden the scope of PCI, has led to the suggestion that which the former were increasingly more aggressive in hybrid revascularization may now equal or even better their percutaneous treatment of coronary lesions (PCI), the results of beating heart surgery. This article reviews and surgeons were developing minimally invasive tech- the strengths and weaknesses of hybrid compared with OPCAB procedures for patients with multivessel disease niques with smaller incisions and performing anastomo- and considers whether DES will lead to a reemergence of ses on the beating heart. It also, and possibly more this technique. importantly, challenged the paradigm that existed in most cardiac surgery centers, that coronary artery bypass grafting (CABG) required cardiopulmonary bypass (CPB) Material and Methods and cardioplegic or other form of arrest. There followed a The MEDLINE and PubMed databases (1996 through resurgence of interest in performing CABG on the beat- March 2004) were searched using the medical subject ing heart, and by the end of the decade off-pump or headings (MeSH) for “coronary artery bypass” combined beating heart CABG (OPCAB) had become an estab- with the following phrases used as text words: “hybrid,” lished technique worldwide. The ability of OPCAB to “integrated,” “MIDCAB” (minimally invasive CABG), “minimally invasive,” “LAST” (left anterior small thora- Accepted for publication July 14, 2004. cotomy), “OPCAB,” “off pump,” “beating heart,” “angioplasty, transluminal, percutaneous, coronary,” and “cor- Address reprint requests to Prof Angelini, Bristol Heart Institute, Bristol Royal Inﬁrmary, Bristol BS2 8HW, UK; e-mail: g.d.angelini@ onary arteriosclerosis” combined with the text words bristol.ac.uk. “drug-eluting stent,” “rapamycin,” “paclitaxel,” “reste-

© 2004 by The Society of Thoracic Surgeons 0003-4975/04/$30.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2004.07.024 1862 REVIEW MURPHY ET AL Ann Thorac Surg REVASCULARIZATION AND DRUG-ELUTING STENTS 2004;78:1861–7 c Abbreviations and Acronyms ASPECT ϭ Asian Paclitaxel-Eluting Stent (%) Survival

Clinical Trial Event-Free BARI ϭ Bypass Angioplasty b

Revascularization Investigation standard deviation.

BHACAS ϭ Beating Heart Against ϭ Redo Cardioplegic Arrest Studies CABG CABG ϭ coronary artery bypass grafting DES ϭ drug-eluting stent MACE ϭ major adverse cardiac event 6–24 0 87% MIDCAB ϭ minimally invasive coronary Mean (Range) 24 (1–60)18 (5–33) 1 1 83% (Months) artery bypass grafting Follow-Up OPCAB ϭ off-pump coronary artery bypass a d d d

grafting Composite endpoint of death, cardiac event, or c (%) PCI ϭ percutaneous coronary TLR intervention RAVEL ϭ randomized study with the sirolimus-eluting, velocity balloon- (%)

expandable stent in the treatment LIMA Patency Immediate of patients with de novo coronary percutaneous coronary intervention; SD

artery lesions ϭ RESEARCH ϭ Rapamycin-Eluting Stent Evaluated at Rotterdam

Cardiology Hospital (%) ϭ

SIRIUS sirolimus-eluting stent Morbidity In-Hospital TAXUS ϭ feasibility study evaluating safety of the NIRx paclitaxel-coated, conformer coronary stent for the treatment of de novo coronary 30-Day

lesions Mortality

TLR ϭ target lesion revascularization Redo CABG required subsequent CABG. b REVIEW target lesion revascularization; PCI

nosis,” “RAVEL” (randomized study with the sirolimus- ϭ

eluting, velocity balloon-expandable stent in the treat- (SD/range) ment of patients with de novo coronary artery lesions), Hospital Stay “SIRIUS” (sirolimus-eluting stent), “ASPECT” (Asian Paclitaxel-Eluting Stent Clinical Trial), “TAXUS” (feasi-

bility study evaluating safety of the NIRx paclitaxel- PCI (Stented) coated, conformer coronary stent for the treatment of de N Vessels novo coronary lesions), “RESEARCH” (Rapamycin- Eluting Stent Evaluated at Rotterdam Cardiology Hospi- tal), or “randomized controlled trial.” In addition, the Vessels reference lists from relevant articles, abstracts, and re- Grafted

views were also searched for additional trials. Random- left internal mammary artery; TLR

ized controlled trials, noncontrolled prospective studies, ϭ and retrospective series, in which morbidity, mortality, (%) angiographic patency, target lesion revascularization Diabetes (TLR), major adverse cardiac events (MACE), or costs Included revision or stenting of LIMA graft. (%)

were reported, were selected for inclusion. Letters and Male d editorials, publications in languages other than English, and duplicate publications were excluded. Individual case reports, pilot studies, and small retrospective stud- Mean ies were not considered when higher levels of evidence range) Age (SD/ were available 1831 63 (35–87)35 69 (48–86) 7850 57 (7) 71 54 (20) 11 83 25 74 18 23 32 42 21 (10) 35 39 (23) 50 5 (1.5) 2.7 (1.0) 47 (14) 50 (39) 7.5 (4) 4.4 (1.7) 0 0 0 0 11 6 100 0 4 100 6% 10% 100 100 18 11 13% 1 1 89% 90% 5442 62 (36–86) 66 (9) 83 76 33 17 63 42 58 49 (36) 3.54 (2–12) 0 0 2 100 12 92 12 (1–23) 12% 0 87% 1437 72 (9)57 66 (8) 93 72 29 11 14 57 28 (20) 37 3.4 72 (2.1) (53) 5.7 (1.8) 40 0 0 21 7 0 100 98 0 14% 1–44 100 0 93% 0–24 92% Results coronary artery bypass grafting; LIMA

Hybrid Revascularization ϭ The results of published series of hybrid procedures are Calculated on a per patient basis (more than one reintervention occurred in some patients). [7] [8] [9] [10] CABG [14] [15] Table 1. Results of Hybrid Revascularization Study (Reference) n a [11] [12] [13] listed in Table 1. These series, although representing a revascularization procedure. Ann Thorac Surg REVIEW MURPHY ET AL 1863 2004;78:1861–7 REVASCULARIZATION AND DRUG-ELUTING STENTS

relatively small patient cohort, demonstrate low proce- whom PCI was not suitable, notably chronically oc- dural morbidity, low mortality, and a short hospital stay cluded, small, calcified, or tortuous vessels or long with this technique. Other advantages are the low fre- atherosclerotic lesions. Although in some hybrid series quency of septic complications and use of homologous completeness of revascularization is as high as 100% [7, blood products, and virtual absence of neurologic injury. 10], in studies with broader inclusion criteria, com- These are attributed to smaller incisions plus avoidance pleteness of revascularization occurred in as few as of CPB and manipulation of the aorta, which have proven 68% of patients [8]. Revascularization of the left ante- to be particularly beneficial in the elderly and in those rior descending coronary artery with the left internal with poor ventricular function, renal impairment, morbid mammary artery is associated with improved survival, obesity, and chronic obstructive airways disease. Twen- freedom from cardiac events, and long-term patency ty-six percent of patients in one study had a Parsonnet compared with PCI and stenting [23–25]. The rationale predicted mortality of 20%, with an overall mortality in for integrating this with PCI to the circumflex and right the series of 0 [8]. Zenati and colleagues [16] compared coronary arteries was that the results of PCI with actual outcome to predicted risk–adjusted mortality in 20 stents, which have early restenosis rates in isolated high-risk patients undergoing minimally invasive coro- short lesions as low as 30%, were comparable to the nary revascularization, including 7 patients who under- long-term survival of saphenous vein grafts [26]. Vein went hybrid procedures. This showed a 30-day mortality grafts, which have a patency of 60% at 10 years [27], are of 0 in a patient cohort with a predicted mortality of 26%. still used in the vast majority of CABG procedures Furthermore, in this cohort the predicted stroke risk, despite the superior long-term patency of arterial calculated using the Multicenter Perioperative Stroke grafts largely owing to their ready availability and the Index, was 22%, with an actual stroke frequency of 0. multivessel nature of the disease [27]. In the “real Despite these excellent early results, hybrid revascu- world,” stents have much higher restenosis rates, how- larization still has a limited role in multivessel revascu- ever, because the pattern of disease is often complex, larization [8, 13]. Reiss and associates [13], in the largest involving bifurcations and sequential and long, diffuse, series of hybrid procedures published to date, performed or calcified lesions, in which reocclusion rates can be as only 57 hybrid procedures during 4 years, compared with high as 60% [28]. This has been shown in randomized 239 OPCAB and 2,305 conventional CABG using CPB. trials in which PCI and stenting is still associated with This represented only 2.2% of the surgical workload. The higher repeat revascularization rates compared with failure of hybrid revascularization to enter widespread OPCAB or conventional CABG [29, 30]. After hybrid use has been attributed to the emergence of OPCAB procedures restenosis of lesions originally treated with surgery [13, 15], which combines many of the benefits of PCI is the most common cause of target vessel failure, minimal invasiveness by avoiding CPB and, when nec- TLR, or MACE, with TLR rates ranging from 12% to essary, aortic manipulation, while at the same time over-

14% at 18 to 24 months’ follow-up (Table 1). In contrast, REVIEW coming many of the limitations of hybrid revasculariza- in patients undergoing OPCAB surgery as part of the tion by permitting unrestricted access to every vessel on the heart and uncompromising selection of the best sites Beating Heart Against Cardioplegic Arrest Studies for coronary anastomoses. Several large series have also (BHACAS) trial, the TLR rate was 2% at a mean highlighted excellent results after OPCAB in those pa- follow-up of more than 2 years (standard deviation, 9 tient groups thought to benefit most from hybrid proce- months) [3]. Similarly, in patients randomized to OP- dures, notably the elderly [17], obese [18], diabetics [19], CAB in the Octopus study [4], TLR rates were 3.5% at or those with poor left ventricular function [20]. The 1 year of follow-up, with 1-year graft patency in a principal benefits of OPCAB compared with hybrid pro- randomized selection of OPCAB patients of 91%. In the cedures relate to the need for repeat revascularization Surgical Management of Arterial Revascularization procedures associated with PCI, the limitations of the Therapies (SMART) randomized trial of off-pump ver- MIDCAB procedure, and the logistic issues associated sus on-pump CABG [6], 1-year angiographic graft with performing two procedures, involving two special- patency after OPCAB surgery was 94%. Hybrid revas- ists, with different procedural protocols, particularly with cularization, unlike OPCAB, has also not been shown respect to anticoagulation. to significantly reduce costs compared with conventional CABG [4–6, 31]. In a randomized trial of hybrid Repeat Revascularization versus conventional CABG, although there was a trend The need for repeat revascularization after hybrid toward a lower initial operative cost in the hybrid procedures is a reflection of the limitations of PCI: high group (€8,100 Ϯ €987 versus €9,700 Ϯ €2,500), this reintervention rates that occur owing to incomplete difference was counterbalanced by the extra cost of revascularization at the time of the primary procedure percutaneous transluminal coronary angioplasty (total as well as post angioplasty or in-stent restenosis [21]. hybrid cost, €10,622 Ϯ €1,329 versus conventional Incomplete revascularization in the Bypass Angio- CABG cost, €9,699 Ϯ €2,500; not significant). At 2 years plasty Revascularization Investigation (BARI) trial, the overall cost in the hybrid group had increased which compared multivessel angioplasty with conven- because of the need for repeat revascularization in 3 tional CABG, occurred in 43% of patients in the PCI patients, although this was not significantly different arm [22] owing to a large subgroup of patients for from the cost of conventional CABG [31]. 1864 REVIEW MURPHY ET AL Ann Thorac Surg REVASCULARIZATION AND DRUG-ELUTING STENTS 2004;78:1861–7

Table 2. Summary of Randomized Trials of Drug-Eluting Stents Versus Non-Drug-Eluting Stents

Mean Mean Vessel Vessel Overall 6-Month 1-Year Length Diameter Diabetics Outcome Outcome n (mm) (mm) (%) (%) TLR MACE Restenosisa (%) TLR MACE Restenosisa

Sirolimus RAVEL [36] 238 9.6 2.6 16 0/26 3/27 0/26 0/26 6/29 0/26 DES/control SIRIUS [37] 1058 14.4 2.8 26 4.1/16.6 4.9/17.7 8.6/21.0 DES/control Paclitaxel TAXUS-1 [38] 61 11.3 3.0 18 0/7 0/10 0/10 3/10 DES/control TAXUS-2 [39] 269 10.4 2.7 12 3.1/14.6 7.8/20.6 5/20 7/19 10/21 MR/control ASPECT [40] 177 10.9 2.9 20 3.3/3.4/3.4 8.3/12.1/16.9 6.6/20.6/45.8 HD/LD/control

a Calculated on a per patient basis (more than one reintervention occurred in some patients). ASPECT ϭ Asian paclitaxel-eluting stent clinical trial; DES ϭ drug-eluting stent; HD ϭ high dose; LD ϭ low dose; MACE ϭ major adverse cardiac event; MR ϭ moderate release; RAVEL ϭ randomized study using the sirolimus-eluting velocity balloon expandable stent for the treatment of patients with de novo coronary artery lesions; SIRIUS ϭ multicenter sirolimus-eluting stent study; TAXUS-1, TAXUS-2 ϭ feasibility study evaluating safety of the NIRx paclitaxel (new intravascular rigid ﬂex) coated conformer, coronary stent for the treatment of de novo coronary lesions; TLR ϭ target lesion revascularization; TVF ϭ target vessel failure.

Morbidity Associated With Minimally Invasive ring as a consequence of PCI complications will increase Coronary Artery Bypass Grafting hospital stay and costs, and reduce the overall benefit Minimally invasive CABG, whether combined with PCI ratio. Performing PCI after surgery permits early angio- or not, also has a procedure-specific morbidity associated graphic assessment of graft patency and protection by REVIEW with the nature of the incisions and the restricted oper- the patent left internal mammary artery graft during PCI, ating field that offsets the benefits of avoiding median and avoids the risk of stent thrombosis in the perioper- sternotomy. Case-controlled cohort studies demonstrate ative period. In the event of a PCI complication or failure, less pain in the days after MIDCAB compared with further emergency CABG may be required, however. OPCAB procedures [32], but report a higher rate of late One way of addressing many of these issues is to perform wound complications in the MIDCAB patients, including PCI and surgery at the same time under general anes- late pain associated with rib retraction and intercostal thetic. This raises logistic issues relating to the workforce neuralgia, and lung herniation [33]. In addition, several planning of two specialists, the cost of developing special studies have reported higher rates of wound infection hybrid operating theaters, and the provision of adequate after MIDCAB [33, 34], attributed to duration of the training. None of these issues are insurmountable, al- procedure and excessive skin retraction. One recent though using two specialists, multiple stents, and spe- study even reported higher rates of graft occlusion, cialized equipment and the risk of restenosis with current reintervention, and MACE after MIDCAB as opposed to PCI would increase cost considerably. The main limiting single-vessel OPCAB [35]. This was thought to be caused factor to performing both procedures at the same time is by a combination of conduit damage during harvest and the need to administer antiplatelet drugs, including in the technical difficulty of performing a beating heart some cases glycoprotein IIb/IIIa inhibitors, to maintain anastomosis with suboptimal access. stent patency, while at the same time performing MIDCAB. Reoperation, either for bleeding after anti- Logistics platelet administration [15], or, as shown in our series [7], Other issues regarding the sequence of, as well as the for stent thrombosis if these agents are not used, again logistics of, performing hybrid procedures remain unre- significantly reduces the cost–benefit ratio. solved. Performing PCI first has the advantage of permitting aggressive multivessel treatment, with the proviso Drug-Eluting Stents that in the event of PCI failure or complication early The DES has been hailed as the answer to in-stent surgical revascularization is planned. The patency of restenosis. Their use, combined with left internal mam- stents may be compromised by the need to discontinue mary artery to left anterior descending coronary artery anticoagulation or antiplatelet therapy for surgery, par- grafting in hybrid procedures in particular, offers the ticularly when ticlopidine, glycoprotein IIb/IIIa inhibi- prospect of minimally invasive revascularization with tors, or even heparin have been administered to treat a excellent long-term outcomes. Summaries of the results complication of PCI. In addition, delay in surgery occur- of several landmark clinical trials of DES versus non-DES Ann Thorac Surg REVIEW MURPHY ET AL 1865 2004;78:1861–7 REVASCULARIZATION AND DRUG-ELUTING STENTS

are listed in Table 2. Sirolimus (rapamycin), a macrolide in diabetics was 17.6% with a target lesion revasculariza- antibiotic with potent immunosuppressive and antipro- tion rate of 6.9%, rising to 10.5% in lesions less than 2.5 liferative effects on vascular smooth muscle cells, was the mm in diameter [37]. Multiple stents, including multives- first agent to be evaluated in clinical trials. The multi- sel stenting, further increase the likelihood of restenosis center, double-blind RAVEL study evaluated the efficacy as a result of the cumulative increase in restenosis risk. of the sirolimus-eluting Bx velocity stent (Cypher; Cordis, There is some evidence from the RESEARCH registry a Johnson & Johnson Company, Miami, FL), and reported that DES are efficacious in high-risk subgroups in the real reductions in stent restenosis from 26.6% to 0% at 210 world [47]. The RESEARCH registry is a single-center days’ follow-up [36]. The larger US multicenter SIRIUS registry that permits comparison of a recent cohort of study also reported a reduction in target vessel failure patients in whom DES have been used without restric- (defined as a composite of deaths from cardiac causes, tion with a historic control group treated before the myocardial infarction, and repeat revascularization) from introduction of DES. They have shown significant reduc- 21% in control patients to 8.6% in the sirolimus group, tions in TLR, 5.1% versus 10.9%, and MACE, 9.7% versus with a TLR rate of 16% and 4% in control patients and 14.8%, in DES and control groups, respectively, at 1 year sirolimus groups, respectively [37]. Follow-up in both the [47]. These benefits were observed despite a higher RAVEL and SIRIUS studies reported almost complete proportion of patients within the DES cohort having abolition of in-stent neointima formation [37, 41, 42]. multivessel disease, multiple stents, bifurcation stenting, Paclitaxel, a taxane with antimitotic properties, has also and type C lesions. It is noteworthy, however, that in a shown remarkable efficacy in clinical trials. Restenosis separate analysis of the same data set de novo in-stent rates of 0% to 4% at 6 and 12 months compared with 11% restenosis within DES was associated with cases that to 27% in control patients have been reported in the were described as “technically complex” [48]. TAXUS 1 (feasibility study evaluating safety of the NIRx Second, animal and clinical studies have suggested paclitaxel coated conformer coronary stent for the treat- that DES may be associated with late restenosis or ment of de novo coronary lesions) [38] and ASPECT thrombosis. Thrombosis has been attributed to the inhi- (Asian Paclitaxel-Eluting Stent Clinical Trial) studies [40]. bition of endothelial regeneration by high local drug Similar advantages in terms of lower TLR rates and concentrations, which is undesirable, in addition to the MACE have been subsequently reported in the larger intended inhibition of neointima formation. This results TAXUS II study [39]. In TAXUS I and II, MACE were in persistently high levels of thrombogenic fibrin and few reduced from 10% to 21% in control patients to 3% to 10% vascular smooth muscle cells around exposed stent struts in the paclitaxel groups at 1 year [38, 39]. In the smaller [49]. An additional problem noted with sirolimus-eluting TAXUS III study, paclitaxel-eluting stents were shown to stents in the RAVEL study [42] was a higher rate of be effective for the treatment of in-stent stenosis [43]. malapposition in patients with eluting stents (21%) ver- The potential for these stents to increase the numbers sus control patients (4%). This may represent expansive of hybrid procedures is limited by several factors. First, remodeling around the stent, possibly related to apopto- REVIEW the applicability of these studies to PCI practice in the sis. Thrombus formation between the stent and vessel real world is unclear. The RAVEL, SIRIUS, TAXUS, and wall may propagate and lead to stent thrombosis. Late ASPECT trials assessed the efficacy of DES in a highly thrombosis leading to myocardial infarction has been selected group of patients, and none published the num- reported with both rapamycin-eluting [49] and paclitax- bers of patients who were initially considered for inclu- el-eluting [50] stents. Late restenosis is also a possibility, sion. Day-to-day clinical practice often involves more- as has been seen after brachytherapy, in which a delayed complex patients than were included in these trials. restenotic response occurs despite impressive early re- There were no implantations in left mainstem disease, or sults [51]. Diminished late efficacy may be caused by a vessels with chronic total arterial occlusion, ostial steno- steady reduction in local tissue levels of the drug. In sis, in-stent restenosis, bifurcation lesions, or bypass sirolimus-eluting stents 63% of the initial dose is eluted grafts, or in patients after acute myocardial infarction by 14 days; at this time arterial wall tissue levels are at a [36]. The effectiveness of DES on completeness of revas- maximum (approximately 160 ␮g) and are reduced 50% cularization and restenosis in these high-risk subgroups by 28 days [52]. A recent study of 15 patients treated for remains to be seen. The stented lesions in these studies in-stent restenosis with QuaDS stents (Quanam Medical were predominantly short lesions in large vessels, fea- Corporation, Santa Clara, CA) containing the paclitaxel tures associated with low restenosis rates with non-DES derivative 7-hexanoyltaxol (QP2 or taxen) has unequivo- [44–46]. Whereas a 2.5- ϫ 20-mm stent would be ex- cally demonstrated late restenosis [53]. Although at 6 pected to have an angiographic restenosis rate of approx- months there was minimal in-stent neointimal hyperpla- imately 32%, a 4- ϫ 8-mm stent would have an angio- sia (late loss, 0.47 Ϯ 1.01 mm), at 12 months there was an graphic restenosis rate of 2% [28]. In the RAVEL study aggressive increase in neointimal growth (late loss, 1.36 only 18% of vessels were less than 2.5 mm, and mean Ϯ 0.94 mm), resulting in a dramatic 61.5% rate of reste- lesion length was small (9.6 mm) [36], with similar obser- nosis. Some series do report continued inhibition of vations in the SIRIUS and TAXUS trials. Another consid- neointima formation at up to 2 years after sirolimus- eration is that diabetics, who constitute 20% of the PCI eluting stent implantation [54, 55]. These were short population, still have higher restenosis rates with DES. In stents (18 mm) placed in large arteries (3 to 3.5 mm) in the SIRIUS study the 6-month in-segment restenosis rate small numbers of low-risk patients, however. 1866 REVIEW MURPHY ET AL Ann Thorac Surg REVASCULARIZATION AND DRUG-ELUTING STENTS 2004;78:1861–7

A final concern is the high cost of stenting multiple 7. Lloyd CL, Calafiore AM, Wilde P, et al. Integrated left vessels with DES. Cypher, the first DES on the market, anterior small thoracotomy and angioplasty for coronary artery revascularization. Ann Thorac Surg 1999;68:908–12. costs approximately five times more than a conventional 8. Zenati M, Cohen HA, Griffith BP. Alternative approach to stent, and would therefore negate any initial cost saving multivessel coronary disease with integrated coronary re- associated with lower TLR rates after PCI, particularly if vascularization. J Thorac Cardiovasc Surg 1999;117:439–44. multiple stents are used to treat several vessels or com- 9. Wittwer T, Cremer J, Boonstra P, et al. Myocardial “hybrid” revascularisation with minimally invasive direct coronary plex lesions. Gunn and colleagues [28] calculated that artery bypass grafting combined with coronary angioplasty: despite some cost savings associated with lower resteno- preliminary results of a multicentre study. Heart 2000;83:58– sis rates, the use of DES would increase overall stent 63. budgets by 256%. 10. Cisowskia M, Morawskia W, Drzewieckib J, et al. Integrated minimally invasive direct coronary artery bypass grafting and angioplasty for coronary artery revascularisation. Eur J Cardiothorac Surg 2002;22:261–5. Conclusions 11. Lewis BS, Porat E, Halon DA, et al. Same-day combined Hybrid procedures currently occupy a select niche in coronary angioplasty and minimally invasive coronary surgery. Am J Cardiol 1999;84:1246–7. particularly high-risk patients or those with favorable 12. Isomura T, Suma H, Hori T, Sato T, Kobashi T, Kanemitsu H. anatomic variants. It remains attractive in situations such Minimally invasive coronary artery revascularization: off- as redo procedures, in patients with significant comor- pump bypass grafting and the hybrid procedure. Ann Tho- rac Surg 2000;70:2017–22. bidity, and in those with severe aortic or mitral ring 13. Riess FC, Bader R, Kremer P, et al. Coronary hybrid revas- calcification in whom moving or elevating the heart cularization from January 1997 to January 2001: a clinical during revascularization of the posterior coronary vessels follow-up. Ann Thorac Surg 2002;73:1849–55. might result in injury of the calcified ascending aorta or 14. Stahl KD, Boyd WD, Vassiliades TA, Karamanoukian HL. Hybrid robotic coronary artery surgery and angioplasty in the mitral annulus. Other patients who may benefit are multivessel coronary artery disease. Ann Thorac Surg 2002; those with prior chest wall irradiation in which median 74(Suppl):S1358–62. sternotomy is relatively contraindicated. In the general- 15. Presbitero P, Nicolini F, Maiello L, et al. “Hybrid” percuta- ity, however, OPCAB permits superior levels of revascu- neous and surgical coronary revascularization: selection criteria from a single-center experience. Ital Heart J 2001;2: larization with lower MACE and fewer TLR and similar 363–8. benefits in terms of low morbidity in high-risk patients. 16. Zenati M, Cohen HA, Holubkov R, et al. Preoperative risk Although undoubtedly effective at reducing in-stent re- models for minimally invasive coronary bypass: a preliminary study. J Thorac Cardiovasc Surg 1998;116:584–9. stenosis, DES will not address the issues of incomplete

REVIEW 17. Koutlas TC, Elbeery JR, Williams JM, Moran JF, Francalancia revascularization or logistic problems associated with NA, Chitwood WR Jr. Myocardial revascularization in the hybrid procedures. Uncertainty regarding the long-term elderly using beating heart coronary artery bypass surgery. effectiveness of DES in many patients as well as their Ann Thorac Surg 2000;69:1042–7. 18. Ascione R, Reeves BC, Rees K, Angelini GD. Effectiveness of high cost, when compared with OPCAB surgery, may coronary artery bypass grafting with or without cardiopul- also limit the wider introduction of hybrid procedures. monary bypass in overweight patients. Circulation 2002;106: 1764–70. 19. Magee MJ, Dewey TM, Acuff T, et al. Inﬂuence of diabetes on References mortality and morbidity: off-pump coronary artery bypass grafting versus coronary artery bypass grafting with cardio- 1. Angelini GD, Wilde P, Salerno TA, Bosco G, Calaﬁore AM. pulmonary bypass. Ann Thorac Surg 2001;72:776–80. Integrated left small thoracotomy and angioplasty for mul- 20. Ascione R, Narayan P, Rogers CA, Lim KH, Capoun R, tivessel coronary artery revascularisation. Lancet 1996;347: Angelini GD. Early and midterm clinical outcome in patients 757–8. with severe left ventricular dysfunction undergoing coro- 2. Hoffman SN, TenBrook JA, Wolf MP, et al. A meta-analysis nary artery surgery. Ann Thorac Surg 2003;76:793–9. of randomized controlled trials comparing coronary artery 21. Pretre R, Turina MI. Choice of revascularization strategy for bypass graft with percutaneous transluminal coronary an- patients with coronary artery disease. JAMA 2001;285:992–4. gioplasty: one- to eight-year outcomes. J Am Coll Cardiol 22. Bypass Angioplasty Revascularization Investigation (BARI) 2003;41:1293–304. Investigators. Comparison of coronary bypass surgery with 3. Angelini GD, Taylor FC, Reeves BC, Ascione R. Early and angioplasty in patients with multivessel disease. N Engl midterm outcome after off-pump and on-pump surgery in J Med 1996;335:217–25. Beating Heart Against Cardioplegic Arrest Studies 23. Diegeler A, Spyrantis N, Matin M, et al. The revival of (BHACAS 1 and 2): a pooled analysis of two randomised surgical treatment for isolated proximal high-grade LAD lesions by minimally invasive coronary artery bypass graft- controlled trials. Lancet 2002;359:1194–9. ing. Eur J Cardiothorac Surg 2000;17:501–4. 4. Nathoe HM, van Dijk D, Jansen EW, et al. A comparison of 24. Drenth DJ, Winter JB, Veeger JGM. Minimally invasive on-pump and off-pump coronary bypass surgery in low-risk coronary artery bypass grafting versus percutaneous trans- patients. N Engl J Med 2003;348:394–402. luminal coronary angioplasty with stenting in isolated high- 5. Ascione R, Lloyd CT, Underwood MJ, Lotto AA, Pitsis AA, grade stenosis of the proximal left anterior descending Angelini GD. Economic outcome of off-pump coronary ar- coronary artery: six months’ angiographic and clinical fol- tery bypass surgery: a prospective randomized study. Ann low-up of a prospective randomized study. J Thorac Cardio- Thorac Surg 1999;68:2237–42. vasc Surg 2002;124:130–5. 6. Puskas JD, Williams WH, Mahoney EM, et al. Off-pump vs 25. Mariani MA, Boonstra PW, Grandjean JG, et al. Minimally conventional coronary artery bypass grafting: early and invasive coronary artery bypass grafting versus coronary 1-year graft patency, cost, and quality-of-life outcomes: a angioplasty for isolated type C stenosis of the left anterior randomized trial. JAMA 2004;291:1841–9. descending artery. J Thorac Cardiovasc Surg 1997;114:434–9. Ann Thorac Surg REVIEW MURPHY ET AL 1867 2004;78:1861–7 REVASCULARIZATION AND DRUG-ELUTING STENTS

26. Feit F, Brooks MM, Sopko G, et al. Long-term clinical Eluting Bx Velocity Balloon-Expandable Stent (RAVEL): outcome in the Bypass Angioplasty Revascularization Inves- sirolimus-eluting stents inhibit restenosis irrespective of the tigation Registry: comparison with the randomized trial. vessel size. Circulation 2002;106:1949–568. BARI Investigators. Circulation 2000;101:2795–802. 42. Serruys PW, Degertekin M, Tanabe K, et al, for the RAVEL 27. Motwani JG, Topol EJ. Aortocoronary vein graft disease: Study Group. Intravascular ultrasound findings in the mul- pathogenesis, predisposition, and prevention. Circulation ticenter, randomized, double-blind RAVEL (RAndomized 1998;97:916–31. study with the sirolimus-eluting VElocity balloon- 28. Gunn J, Morton AC, Wales C, Newman CM, Crossman DC, expandable stent in the treatment of patients with de novo Cumberland DC. Drug eluting stents: maximising benefit native coronary artery Lesions) trial. Circulation 2002;106: and minimising cost. Heart 2003;89:127–31. 798–803. 29. Serruys P, Unger F, Sousa J, et al. Comparison of coronary 43. Tanabe K, Serruys PW, Grube E, et al. In-stent restenosis artery bypass surgery and stenting for the treatment of with stent based delivery of paclitaxel incorporated in a multivessel disease. N Engl J Med 2001;344:1117–24. slow-release polymer formulation. Circulation 2003;107:559– 30. Eefting F, Nathoe H, van Dijk D, et al. Randomized compar- 64. ison between stenting and off-pump bypass surgery in 44. Serruys PW, Kay P, Disco C, et al. Periprocedural quantita- patients referred for angioplasty. Circulation 2003;108: tive coronary angiography after Palmaz-Schatz stent im- 2870–6. plantation predicts the restenosis rate at six months. J Am 31. de Canniere D, Jansens JL, Goldschmidt-Clermont P, et al. Combination of minimally invasive coronary bypass and Coll Cardiol 1999;34:1067–74. percutaneous transluminal coronary angioplasty in the 45. Kobayashi Y, De Gregorio J, Kobayashi N, et al. Stented treatment of double-vessel coronary disease: two-year fol- segment length as an independent predictor of restenosis. low-up of a new hybrid procedure compared with “on- J Am Coll Cardiol 1999;34:651–9. pump” double bypass grafting. Am Heart J 2001;142:563–70. 46. Kornowski R, Bharghava B, Fuchs S, et al. Procedural results 32. Gulielmos V, Brandt M, Dill HM, et al. Coronary artery and late clinical outcomes after percutaneous interventions Ն Ͻ bypass grafting via median sternotomy or lateral minithora- using long ( 25 mm) versus short ( 20 mm) stents. J Am cotomy. Eur J Cardiothorac Surg 1999;16(Suppl):S48–52. Coll Cardiol 2000;35:612–8. 33. Gersbach P, Imsand C, von Segesser LK, Delabays A, Vogt P, 47. Lemos PA, Serruys PW, van Domburg RT, et al. Unrestricted Stumpe F. Beating heart coronary surgery: is sternotomy a utilization of sirolimus-eluting stents compared with con- suitable alternative to minimal invasive technique. Eur ventional bare stent implantation in the “real world”: the J Cardiothorac Surg 2001;20:760–4. Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiol- 34. Detter C, Reichenspurner H, Boehm D, Thalhammer M, ogy Hospital (RESEARCH) registry. Circulation 2004;109: Schutz A, Reichart B. Single vessel revascularisation with 190–5. beating heart techniques—minithoracotomy or sternotomy? 48. Lemos PA, Saia F, Ligthart JM, et al. Coronary restenosis Eur J Cardiothorac Surg 2001;19:464–70. after sirolimus-eluting stent implantation: morphological 35. Vicol C, Nollert G, Mair H, et al. Midterm results of beating description and mechanistic analysis from a consecutive heart surgery in 1-vessel disease: minimally invasive direct series of cases. Circulation 2003;108:257–60. coronary artery bypass versus off-pump coronary artery 49. Choi SB. CYPHER coronary stents and risk of thrombosis. bypass with full sternotomy. Heart Surg Forum 2003;6:341–4. Can Med Assoc J 2003;169:218. 36. Morice MC, Serruys PW, Sousa JE, et al, for the RAVEL 50. Liistro F, Stankovic G, Di Mario C, et al. First clinical Study Group. Randomized Study with the Sirolimus-Coated experience with a paclitaxel derivate-eluting polymer stent Bx Velocity Balloon-Expandable Stent in the Treatment of system implantation for in-stent restenosis: immediate and REVIEW Patients with de Novo Native Coronary Artery Lesions. A long-term clinical and angiographic outcome. Circulation randomized comparison of a sirolimus-eluting stent with a 2002;105:1883–6. standard stent for coronary revascularization. N Engl J Med 51. Kay IP, Wardeh AJ, Kozuma K, et al. Radioactive stents delay 2002;346:1773–80. but do not prevent in-stent neointimal hyperplasia. Circula- 37. Moses JW, Leon MB, Fitzgerald PJ, et al, for the SIRIUS tion 2001;103:14–7. Investigators. Sirolimus-eluting stents versus standard 52. Klugherz BD, Llanos G, Lieuallen W, et al. Twenty-eight-day stents in patients with stenosis in a native coronary artery. efficacy and pharmacokinetics of the sirolimus-eluting stent. N Engl J Med 2003;349:1315–23. 38. Grube E, Silber S, Hauptmann KE, et al. TAXUS I: six- and Coron Artery Dis 2002;13:183–8. twelve-month results from a randomized, double-blind trial 53. Virmani R, Liistro F, Stankovic G, et al. Mechanism of late on a slow-release paclitaxel-eluting stent for de novo coro- in-stent restenosis after implantation of a paclitaxel deri- nary lesions. Circulation 2003;107:38–42. vate-eluting polymer stent system in humans. Circulation 39. Colombo A, Drzewiecki J, Banning A, et al. TAXUS-II Study 2002;106:2649–51. Group. Randomized study to assess the effectiveness of 54. Sousa JE, Costa MA, Sousa JG, et al. Two-year angiographic slow- and moderate-release polymer-based paclitaxel- and intravascular ultrasound follow-up after implantation of eluting stents for coronary artery lesions. Circulation 2003; sirolimus-eluting stents in human coronary arteries. Circu- 108:788–94. lation 2003;107:381–3. 40. Park SJ, Shim WH, Ho DS, et al. A paclitaxel-eluting stent for 55. Degertekin M, Serruys PW, Foley DP, et al. Persistent the prevention of coronary restenosis. N Engl J Med 2003; inhibition of neointimal hyperplasia after sirolimus-eluting 348:1537–45. stent implantation: long-term (up to 2 years) clinical, angio- 41. Regar E, Serruys PW, Bode C, et al. Angiographic findings of graphic, and intravascular ultrasound follow-up. Circulation the multicenter Randomized Study With the Sirolimus- 2002;106:1610–3.