Complete Versus Incomplete Revascularization With Coronary Artery Bypass Graft or Percutaneous Intervention in Stable Coronary Artery Disease
2012; Lippincott Williams & Wilkins; Volume: 5; Issue: 4 Linguagem: Inglês
10.1161/circinterventions.111.965509
ISSN1941-7632
AutoresMario Gössl, David P. Faxon, Malcolm R. Bell, David R. Holmes, Bernard J. Gersh,
Tópico(s)Cardiac and Coronary Surgery Techniques
ResumoHomeCirculation: Cardiovascular InterventionsVol. 5, No. 4Complete Versus Incomplete Revascularization With Coronary Artery Bypass Graft or Percutaneous Intervention in Stable Coronary Artery Disease Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBComplete Versus Incomplete Revascularization With Coronary Artery Bypass Graft or Percutaneous Intervention in Stable Coronary Artery Disease Mario Gössl, MD, David P. Faxon, MD, Malcolm R. Bell, MBBS, David R. Holmes, MD and Bernard J. Gersh, MB ChB, DPhil Mario GösslMario Gössl From the Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota (M.G., M.R.B., D.R.H., B.J.G.); Department of Medicine, Division of Cardiology, Brigham and Women's Hospital, Boston, Massachusetts (D.P.F.) , David P. FaxonDavid P. Faxon From the Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota (M.G., M.R.B., D.R.H., B.J.G.); Department of Medicine, Division of Cardiology, Brigham and Women's Hospital, Boston, Massachusetts (D.P.F.) , Malcolm R. BellMalcolm R. Bell From the Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota (M.G., M.R.B., D.R.H., B.J.G.); Department of Medicine, Division of Cardiology, Brigham and Women's Hospital, Boston, Massachusetts (D.P.F.) , David R. HolmesDavid R. Holmes From the Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota (M.G., M.R.B., D.R.H., B.J.G.); Department of Medicine, Division of Cardiology, Brigham and Women's Hospital, Boston, Massachusetts (D.P.F.) and Bernard J. GershBernard J. Gersh From the Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota (M.G., M.R.B., D.R.H., B.J.G.); Department of Medicine, Division of Cardiology, Brigham and Women's Hospital, Boston, Massachusetts (D.P.F.) Originally published1 Aug 2012https://doi.org/10.1161/CIRCINTERVENTIONS.111.965509Circulation: Cardiovascular Interventions. 2012;5:597–604IntroductionIn patients who have multivessel but stable coronary artery disease (CAD), the perceived advantages of complete revascularization (CR) over incomplete revascularization (IR) are intuitively logical and have been the impetus to perform CR whenever feasible. These advantages were suggested by early studies in patients with 3-vessel CAD who underwent coronary artery bypass graft (CABG) demonstrating a survival benefit of patients with CR compared with those with IR.1–4 Nonetheless, in practice, the variations in severity of the coronary anatomy as well as the patient's clinical status often precludes CR resulting in a higher frequency of IR in patients with multivessel disease. For example, in the SYNTAX (SYNergy between PCI with TAXus and Cardiac Surgery) trial, patients randomized to percutaneous intervention (PCI)5 had CR in only 56.7% even when CR was intended and in Bypass Angioplasty Revascularization Investigation (BARI) this was 56%.6 A critical review of the literature demonstrates the complexity of the issues and poses several unanswered questions (Table 1).Table 1. Unanswered QuestionsUnanswered Questions1. Are there standardized definitions for CR/IR available?2. Is CR a fundamental tenet or is it just a worthwhile objective, for which benefits outweigh the risks? Does it have the same implications for surgeons vs interventional cardiologists3. Should CR become the standard for comparison of the efficacy of different procedures, eg, should the ability to achieve CR vs IR be used as a criterion to select specific therapeutic options such as PCI vs CABG?4. Do we perform CR in those patients in whom we can, –and only perform IR when CR is not feasible?5. Has the FAME7 study reframed the issues with regard to CR vs IR?6. Does the effect of CR vs IR depend on the specific arterial segment involved, eg, is CR more important when the LAD is involved?CR indicates complete revascularization; IR, incomplete revascularization; PCI, percutaneous intervention; FAME, Fractional Flow Reserve vs Angiography for Multivessel Evaluation; and LAD, left anterior descending coronary artery.In the current era of significantly improved medical therapy (primary and secondary prevention) of CAD as well as the increasing use of sophisticated invasive and noninvasive testing of the physiological significance of coronary artery stenoses, both surgeons and cardiologists have an arsenal of diagnostic tools at hand that may guide decisions in selecting revascularization strategies. This review provides a contemporary overview of the currently available literature in the field and proposes an evidence-based approach for patients with severe CAD and stable angina undergoing revascularization.Frequently Encountered Difficulties in Reviewing Data on CRThe Missing Universal DefinitionThe dilemma of comparing outcomes after CR and IR begins with the lack of universal definitions. There are considerable differences between complete/incomplete anatomic and functional/physiological revascularization. Ong and Serruys8 reviewed comprehensively past definitions of CR used in different trials (Table 2).Table 2. Summary of Definitions for Complete Revascularization After Ong and Serruys et al8RevascularizationDefinitionComplete anatomic/numeric revascularization•Unconditional3,9,10•Conditional11–13All stenotic vessels are revascularized, irrespective of size and territory supplied.All stenotic vessels greater than a defined diameter are revascularized, ORAll stenotic main-branch vessels are revascularized.Complete functional revascularization14,15All ischemic myocardial territories are reperfused; areas of old infarction with no viable myocardium are not required to be reperfused.Complete revascularization by a predetermined scoring cutoff value16Scoring of stenoses in different vessels at different locations (weightings may be used). The overall extent of disease is a continuous variable, the treatment is another variable, and the posttreatment score determines completeness of revascularization.Functional17Jeopardy score: The postrevascularization score is calculated on the basis of the amount of remaining myocardium at risk.As seen, some definitions are based on anatomy, some on the extent of ischemia, and some on other preset criteria. Conceivably, a comparison of studies using such different definitions is difficult, if not impossible. In most trials, completeness of revascularization has been determined on anatomic basis because of lack of data on myocardial viability and jeopardy.18 To improve future comparability, Zimarino et al18 recently proposed a contemporary definition of 3 different types of revascularization as follows: (1) complete anatomic revascularization, defined as treatment of all coronary artery segments >1.5 mm in diameter and ≥50% diameter stenosis; (2) incomplete anatomic but functionally adequate revascularization (ie, reasonable IR or functionally CR),19 defined as treatment of all coronary segments with ≥50% diameter stenosis supplying viable myocardium; and (3) incomplete functional revascularization consequentially, defined as the inability to treat all coronary segments that supply viable myocardium and have a >50% diameter stenosis.18Influence of Confounding VariablesIn both trials and registries, there are multiple reasons underlying the decision not to perform CR in an individual patient and differences in baseline variables appear to play a major confounding role, which could bias the data in favor of CR by selecting the healthier patient for CR rather than IR. Multivariate analyses tend to adjust for differences in baseline variables but cannot eliminate these nor adjust for unmeasured confounders/variables and no amount of propensity matching can evaporate the bias introduced when one group of patients is sicker than the other.The clinical presentation influences the decision to perform CR; for example, the presence of hemodynamic instability at the time of the procedure may have an impact on procedural performance, in both CABG and PCI.20Other patient characteristics also influence the decision such as older age, comorbidities, and the development of a procedural complication. In general, sicker patients with multiple, significant comorbidities (including left ventricular systolic failure, prior myocardial infarctions, peripheral artery disease, renal failure, and diabetes mellitus) frequently have IR as demonstrated by several PCI registries.21–24 In addition, anatomic differences, which include chronic total occlusion ([CTO], probably the most important reason for IR with PCI), small vessel disease, diffuse atherosclerosis, the extent of myocardial jeopardy, and myocardial viability, may strongly influence the decision regarding the extent of revascularization.Determination of the success of CR is often based on the immediate procedural outcome. This ignores, however, the contribution of early or late bypass graft failure and early or late in-stent restenosis; nevertheless, both are incorporated into the end point of repeat revascularization. Moreover, PCI trials often have an inherent methodological disadvantage. Procedural success is assessed with postprocedural coronary angiography, whereas for CABG trials the review of completeness of revascularization often relies on the operative report to determine (1) which vessels were bypassed and (2) the status of early graft patency (assessed by flowmeters and occasionally fluorescein angiography). In addition, angiographic evaluations often lack the use of quantitative coronary angiography, which significantly increases inter- and intraobserver variability.IR Versus CR in Chronic Stable Angina: Trials, Registries, and Single-Center ExperiencesCR Versus IR in Coronary Artery Bypass Grafting Alone or In Direct Comparison With PCIThere are several studies that suggest a survival benefit of CR over IR in patients with severe multivessel CAD undergoing CABG. A large study from Emory showed that survival at 5 years was significantly greater in patients with CR (88.5%) than in those with IR (83.5%).3 In addition, more patients were free of angina after CR (70%) than after IR (58%). Similarly, Kleisli et al11 demonstrated that CR was associated with better survival (5-year unadjusted survival rate 82.4% versus 52.6%), only limited by major baseline differences between the 2 groups favoring patients who underwent CR, and a lack of adjustment in the survival analysis. Another large series from the Cleveland Clinic showed that CR compared with IR with ungrafted high-grade left circumflex or right CAD was associated with a substantially increased 10- and 20-year survival (91.1% CR versus 81% IR at 10 years and 70% versus 53% at 20 years).25 Finally, a more recent substudy of the SYNTAX trial including registry and randomized data demonstrated, after multivariate analysis, that IR and not the complexity of the coronary anatomy to be an independent predictor of adverse 2-year outcomes (P=0.002).12 The difference, however, was mainly driven by the need for repeat revascularization and not by survival (Tables 3 and 4).Table 3. Summary Data From Studies of CABGTrial/Study/RegistryYear(s)No. of Patients/Follow-UpDefinition of CR7End PointsBenefit of CR vs IR/CaveatsStudy DesignCASS, Bell91974–19793372 with 3-vessel disease/annually for a mean of 4.9 y (maximum 8.1)Unconditional (≥3 vessels bypassed)Freedom from death, myocardial infarction, reoperation, or development of definite anginaOnly in CCS III/IV (in particular, those with EF 70%) LCX, prox. RCA, and any left main disease reduced long-term survivalObservationalBARI, Vander Salm271988–19911507/7 yMultiple (see Table 2)Death, cardiac death, MI, repeat revascularization, angina, composites of death/MI, and cardiac death/MINone; multiple non-LAD bypass grafts may increase riskObservationalKleisli111998–20001034/5 yConditional (grafts to every primary coronary artery >50% stenosis)All-cause death and cardiac deathSurvival benefit of CR over IR/unadjusted Kaplan-Meier curves only despite significant baseline differencesObservationalRastan282000–2007936 (of 8806)/1 and 5 yReasonable IR (LIMA to LAD only)Cumulative survival and hospital mortalityNo differenceObservationalMohr122005–20071541/2 yConditional, grafting all lesions ≥50% with a vessel diameter >1.5 mmMACCEs = all-cause death, VCA, MI, and repeat revascularizationIR (driven by repeat revascularization) predictor for adverse 2-y outcomeRCT/RegistryKozower291986–2003500/5 and 8 yConditional (≥1 graft to all major territories with ≥50% stenosis)SurvivalCR improved mean survival by 25% (only patients 80–94 y old)ObservationalAziz301986–2007580/5 and 8 yUnconditional and conditional (≥1 graft to all or major diseased vessels ≥50% stenosis)SurvivalCR improved mean survival by 18% (only patients 80–94 y old)ObservationalCABG indicates Coronary Artery Bypass Graft; CR, complete revascularization; IR, incomplete revascularization; CASS, Coronary Artery Surgery Study; CSS, Canadian Cardiovascular Society; LAD, left anterior descending coronary artery; RCA, right coronary artery; LCX, left circumflex; BARI, Bypass Angioplasty Revascularization Investigation; and MI, myocardial infarction.Table 4. Data From Subset Analyses From PCI vs CABG TrialsTrial/Study/RegistryYear(s)No. of patients/Follow-UpDefinition of CR8/End PointsBenefit of CR vs IR/CaveatsBARI, Bourassa141988–19911829/5 yFunctional/mortality, cardiac mortality, repeat CABG, repeat angioplasty, and any repeat revascularizationPTCA-IR in nondiabetics does not compromise long-term survivalARTS I101997–19981143/1 yUnconditional (all lesions with ≥50% stenosis)/death, MI, and CVAIR-PCI with reduced event-free survival/only with inclusion of repeat revascularizationARTS II13567/5 yConditional (all lesions with ≥50% stenosis ≥1.5 mm)/death, CVA, MI, and any revascularizationIR-PCI confers higher event rate only in highest SYNTAX score tertile/major driver of statistics was repeat revascularizationKim312003–20051914/5 yMultiple/death, composite of death, MI, CVA, and composite of death, MI, CVA, and repeat revascularizationNo clear benefit of CR, only trend toward increased composite of death, MI, CVA, and repeat revascularization with multivessel (≥2) IR-PCIPCI indicates percutaneous intervention; CABG, coronary artery bypass graft; CR, complete revascularization; IR, incomplete revascularization; BARI, Bypass Angioplasty Revascularization Investigation; PTCA, percutaneous transluminal coronary angioplasty; ARTS, Arterial Revascularization Therapies Study; MI, myocardial infarction; and CVA, cerebrovascular accident.In contrast to the above, several other studies failed to demonstrate convincing benefit of CR over IR for multivessel CABG. In an important but retrospective study, using the CASS Registry data, Bell et al9 showed that in stable patients (group I, Canadian Cardiovascular Society I–II), the number of vessels bypassed failed to significantly influence event-free survival. In patients who had severe symptoms (Canadian Cardiovascular Society classes III–IV [group II]), however, bypassing ≥3 vessels independently improved long-term survival, specifically in patients with left ventricular dysfunction (ejection fraction 1 graft to right coronary artery, circumflex, or ramus system). In addition, the subsequent subanalysis of Bourassa et al14 within the nondiabetic group with a priori intended IR failed to show a difference in the 5-year overall and cardiac mortality. More data supporting noninferiority of IR were recently published by Rastan et al.28 In this analysis of 936 consecutive patients undergoing reasonable IR (with a left internal mammary artery to the LAD but without bypasses to the right coronary artery or circumflex vessels), there was no difference in hospital mortality (3.3% CR versus 3.2% IR) or in cumulative survival at 1 year (93.1% CR versus 93.6% IR) and 5 years (82.2 CR versus 80.9% IR). Further indirect support for the concept of reasonable IR comes from older data by Lytle et al,26 where reoperation of a late graft stenosis (≥5 years) conveyed no survival benefit, unless performed on an LAD vein graft, in patients with moderate-to-severe decrease in LV function, older age, 3 vessels or left main disease, and classes I and II symptoms.Download figureDownload PowerPointFigure 1. Selected graphs from Bell et al9 showing event-free survival stratified by several vessels grafted overall (upper left graphic), survival in patients with an ejection fraction (EF) 2 decades. Within this time frame, interventional techniques have significantly and rapidly improved including the implementation of bare metal and subsequently DES over balloon angioplasty alone (Table 5).Table 5. Summary Data From Studies of PCITrial/Study/RegistryYear(s)No. of patients/Follow-UpDefinition of CR8/End PointsBenefit of CR vs IR/CaveatsNHLBI PTCA, Bourassa231985–1986757/9 yResidual stenosis <50%/freedom from death, MI, CABG, and PTCACR reduced late occurrence of CABG onlyBell321979–1988867/26 moDilation of all stenoses ≥70%/event-free survival of death, MI, CABG, and anginaNo difference after adjustment for baseline variablesBARI, Kip331988–19912047/5 yAll clinically relevant lesions/cardiac death, death/MI, and CABGCR reduced the need for CABGIjsselmuiden341995–1998219/1 and 5 yAll vessels ≥50% stenosis/composite of cardiac and noncardiac death, CABG, and re-PCINo differenceNYS221997–200021 945/3 yAll lesions ≥50% in major vessels/survivalIR increased adjusted mortality by 15% (absolute survival differences small, baseline characteristics favoring CR)/CTO patient populationAPPROACH151995–20011956/3 yFunctional (Duke jeopardy score 2 mm)/cardiac survival10% survival benefit with CR/CTO patient populationLehmann362000–2008679/2.5 y3 different definitions, unconditional and conditional/long-term all-cause mortalityOnly CR of all segments (unconditional) predictor of survival/IR patients with more comorbiditiesGéneréux162003–20052686/1 yResidual SYNTAX score=0/MACE and death or MI after 30 days and 1 yResidual SYNTAX score >8 associated with poor prognosis/ACS patient populationPCI indicates percutaneous intervention; CR, complete revascularization; IR, incomplete revascularization; MI, myocardial infarction; CABG, coronary artery bypass graft; PTCA, percutaneous transluminal coronary angioplasty; CTO, chronic total occlusion; and APPROACH, Alberta Provincial Project for Outcomes Assessment in Coronary Heart Disease.This has had major implications for the comparability and applicability of studies performed before the era of routine use of DES and current improvements in postinterventional medical therapy, including more advanced antiplatelet therapy and aggressive risk factor reduction.Several earlier studies and registries, some reaching into the BMS era, failed to show a clear long-term survival benefit of CR over IR in percutaneous interventions. Bourassa et al23 investigated the long-term outcomes of 757 patients in the National Heart, Lung, and Blood Institute (NHLBI) Percutaneous Transluminal Coronary Angioplasty (PTCA) Registry, 83% of who had IR. In-hospital emergency and elective CABG rates were higher for IR; the 9-year data, however, showed no significant difference in adjusted freedom from death, MI, recurrent angina, or repeat revascularization between CR and IR. Data from Mayo Clinic in the early PTCA era demonstrated that CR was achieved in only 41%.32 After 26 months of follow-up, adjusted event-free survival of death or MI, necessity for CABG surgery, and the occurrence of severe angina were similar between IR and CR.Ijsselmuiden et al34 followed 219 patients with MVD in the bare-metal stent era of which 108 underwent CR and 111 culprit vessel PCI (CVR). Within a 1- and 5-year follow-up, there was no statistically significant difference in MACE between CVR and CR. Data from the BARI trial showed that within 5 years after PTCA, there was no significant difference between patients receiving CR versus IR with angioplasty with regard to cardiac death, MI, need for CABG, or angina.14 By pooling data from 2047 patients undergoing PTCA in the BARI randomized trial and the BARI observational registry, Kip et al33 investigated the differential impact of a pre-PTCA strategy of IR versus IR because of initial lesion outcome. Despite unfavorable baseline characteristics of the IR group, a pre-PTCA strategy of IR or IR because of unsuccessful but intended CR was unrelated to 5-year risk of cardiac death or death/myocardial infarction compared with CR, but were all independently related to the need for CABG.33 In contrast, more recent data in the later BMS and DES era, especially in patients with CTO, suggest that CR has a small but statistically significant survival benefit over IR.Data from the New York State Registry during the BMS era tend to support the concept of CR by demonstrating that the 3-year adjusted mortality was increased by 15% when revascularization was incomplete.22 In detailed analyses of the IR group, a single unopened CTO increased the risk of late death by 35% and ≥2 incompletely revascularized arteries including ≥1 CTO increased the risk by 36%. Despite statistical significance, the magnitude of the benefit was small and of uncertain clinical significance. In addition, there were multiple unfavorable patient characteristics and a variety of comorbidities associated with IR that all favored the CR arm reflecting the limitations of multivariate analysis as eluded to above.Comparable data with similarly CR-favoring baseline characteristics were obtained by the same group during the DES era.21 In this large cohort, 18-month survival was slightly, but still statistically significantly higher in the CR group (93.8 versus 94.9%; P=0.01). When stratified further, patients with 2 vessels not revascularized and total occlusions were at highest risk.Additional support for performing CR in severe CAD with at least 1 CTO came from Valenti et al.35 The 2-year cardiac survival rate was higher in the CTO-PCI success group compared with CTO-PCI failure group and overall in patients with CR when compared with patients with IR.The investigators of the Alberta Provincial Project for Outcomes Assessment in Coronary Heart Disease (APPROACH)15 reported that independent predictors of IR were the presence of a total occlusion, a higher pre-PCI Duke jeopardy score,17 aged >65 years, and renal failure. After correcting for baseline characteristics, complete multivessel PCI was associated with a reduced need for future CABG but not repeat PCI and only a nonsignificant trend toward better survival.More recently, Lehmann et al36 demonstrated that CR was associated with a survival benefit of ≈50% over IR. When STEMI patients (26% of the population) were excluded, however, there was no longer a statistically significant difference.In a meta-analysis of 26 studies and a total of 46 260 patients by Bangalore et al,37 there was an overall trend of favoring CR. CR was associated with a 35% lower risk of all-cause death, significantly lower cardiac mortality, nonfatal myocardial infarction (OR=0.79; 95% CI, 0.70–0.89) and the need for CABG (OR=0.51; 95% CI, 0.39–0.65) compared with patients undergoing IR. CR was also associated with 31% less angina with no difference in repeat PCI. The results were mainly driven by the BARI and NYS-PCI registry data.37Most recently, a novel approach of calculating the residual SYNTAX score in patients with IR after PCI for moderate- to high-risk acute coronary syndromes demonstrated a poorer 30-day and 1-year prognosis for patients with a residual SYNTAX score of >8.16 Future studies in patients with stable CAD will have to show whether a higher residual SYNTAX score is also associated with a poorer prognosis in a stable patient population.In conclusion, the sicker the patient as defined by the presence of multivessel disease, CTO and impaired LV function combined with the evidence of viable myocardium, the greater the apparent survival benefit from CR with PCI but with the caveat that in almost all the studies baseline variables and demographics tended to favor patients receiving CR. In addition, only ≈50% of patients could be completely revascularized largely because of the presence of CTO. In studies where CR is obtainable, outcomes are better when it is achieved. The actual differences in survival are small, and the major benefit may be on a lower rate of repeat revascularization in the DES era. Irrespective of whether the patient is completely or incompletely revascularized, optimal medical therap
Referência(s)