Skip to content

Revascularization — PCI and CABG

TL;DR — Revascularization is not one intervention: PCI and CABG differ in invasiveness, completeness, durability, peri-procedural hazards, and downstream events. In stable disease without left-main or other high-risk anatomy, routine PCI does not reduce death or MI over medical therapy, although it can relieve concordant angina (Maron 2020, PMID 32227755; Rajkumar 2023, PMID 38015442). For complex multivessel disease—especially diabetes—CABG generally trades more early surgical hazard for fewer spontaneous MI and repeat procedures; an individual-patient meta-analysis found lower 5-year mortality with CABG, with effect modified by diabetes and complexity (Head 2018, PMID 29478841). Left-main trials support PCI as an alternative in selected low-complexity anatomy, but endpoint definitions and longer-term spontaneous MI/reintervention differences remain contested (Stone 2016, PMID 27797291; Sabatine 2021, PMID 34793745). In ischemic LV dysfunction, 10-year STICHES supports CABG, whereas REVIVED found no death/HF-hospitalization benefit from PCI despite viability (Velazquez 2016, PMID 27040723; Perera 2022, PMID 36027563).

What each strategy changes

Strategy Immediate target Strength Main limitation
PCI One or more stenoses Low initial invasiveness; rapid symptom relief Restenosis, thrombosis, repeat procedures; focal treatment
CABG Flow beyond proximal disease using grafts More complete/durable protection in diffuse complex disease Sternotomy, stroke, infection, recovery, graft failure
Medical therapy Systemic plaque/thrombotic biology and demand Treats entire arterial bed; no procedural harm Adherence; may not control symptoms or high-risk anatomy

The comparison must specify the clinical syndrome, anatomy, LV function, diabetes, operative risk, symptom burden, feasibility of complete revascularization, and patient preference.

Stable disease: no routine prognostic PCI

COURAGE and ISCHEMIA found no reduction in death or MI from routine PCI/invasive management added to medical therapy in trial-eligible stable patients (Boden 2007, PMID 17387127; Maron 2020, PMID 32227755). ISCHEMIA-EXTEND found no all-cause survival advantage at approximately 7 years despite fewer cardiovascular and more non-cardiovascular deaths in the invasive group (Hochman 2023, PMID 36335918).

ORBITA-2 establishes a symptom indication: PCI improved angina compared with a placebo procedure (Rajkumar 2023, PMID 38015442). The appropriate inference is therefore “symptom-directed PCI after concordance,” not “PCI is ineffective.”

PCI versus CABG by anatomy

Anatomy/phenotype Evidence signal Typical decision emphasis
Focal low-complexity disease PCI often achieves complete treatment with short recovery Symptoms, DAPT tolerance, restenosis risk
Left main, low complexity PCI and CABG both feasible in selected patients Early recovery versus later spontaneous MI/reintervention
Three-vessel/high SYNTAX complexity CABG generally favored Completeness and durability
Diabetes plus multivessel disease CABG benefit more pronounced in pooled evidence Survival/MI versus surgical hazard
High surgical risk/frailty PCI may be the only practical revascularization route Expected survival, symptom goal, procedural futility

At 10 years in SYNTAX, outcomes differed by three-vessel versus left-main strata, reinforcing that “multivessel/left main” should not be treated as one anatomic category (Thuijs 2019, PMID 31488373). The pooled individual-patient analysis of randomized trials reported lower 5-year all-cause mortality with CABG than PCI overall, with larger differences in diabetes and greater anatomic complexity (Head 2018, PMID 29478841).

For left-main disease, EXCEL showed comparable early composite results with everolimus-eluting stents and CABG in selected low/intermediate-complexity anatomy, while later outcomes and definitions generated sustained controversy (Stone 2016, PMID 27797291; Azzalini 2020, PMID 32921366). An individual-patient meta-analysis supports no large difference in 5-year mortality but more spontaneous MI and repeat revascularization with PCI and more procedural MI with CABG depending on definition (Sabatine 2021, PMID 34793745). The 2022 ESC/EACTS review documents how evidence interpretation affected guideline recommendations (Byrne 2023, PMID 37632766).

Diabetes

Diabetes is a marker of diffuse disease, smaller vessels, accelerated atherosclerosis, and graft/PCI risk. In pooled randomized evidence, the mortality advantage of CABG over PCI was concentrated in patients with diabetes rather than those without it (Head 2018, PMID 29478841).

Decision-making still requires operative risk, conduit quality, kidney disease, frailty, ability to complete revascularization, and expected lifespan. “Diabetes means CABG” is a population-level summary, not an automatic bedside conclusion.

Ischemic LV dysfunction: CABG and PCI are not interchangeable

Trial Procedure Population Main result
STICHES CABG + medical therapy vs medical therapy EF ≤35%, CABG-suitable disease Lower all-cause death over 10 years with CABG (Velazquez 2016, PMID 27040723)
REVIVED-BCIS2 PCI + medical therapy vs medical therapy EF ≤35%, viable myocardium, PCI-suitable disease Death/HF hospitalization 37.2% vs 38.0%; HR 0.99 (Perera 2022, PMID 36027563)

REVIVED also found no meaningful recovery of LV function from PCI, challenging the assumption that viability identifies myocardium whose prognosis improves after PCI (Perera 2022, PMID 36027563). STICHES supports CABG over a longer horizon but included operative hazard and a surgically suitable population (Velazquez 2016, PMID 27040723).

Cross-trial comparison cannot prove that CABG is superior to PCI in the same patient; route, completeness, anatomy, selection, and follow-up differ. STICH3C is directly testing CABG versus PCI in ischemic cardiomyopathy (NCT05427370, live ClinicalTrials.gov query re-run 2026-09-02).

Physiology guidance

FFR and iFR estimate whether a stenosis causes a pressure drop under defined conditions. They improve lesion selection compared with angiography alone, but discordance occurs and long-term pooled analyses have raised questions about equivalence on mortality (Eftekhari 2023, PMID 37634144; Berry 2017, PMID 28972006).

Tool Measurement Strength Limitation
FFR Hyperemic distal/aortic pressure ratio Large evidence base; lesion-specific Adenosine; microvascular state affects result
iFR Resting wave-free pressure ratio No hyperemic drug Discordance near thresholds
IVUS Lumen, plaque burden, stent expansion Strong for left main and optimization No direct ischemia measure
OCT High-resolution cap, thrombus, stent detail Mechanism and optimization Contrast load, limited depth
CCTA/FFR-CT Noninvasive anatomy/estimated physiology Whole-coronary map Image quality and model dependence

Placebo-controlled ORBITA analyses show that lower FFR/iFR predicts larger symptom benefit from PCI, not established survival benefit (Foley 2025, PMID 39462291).

Complete revascularization in ACS

In STEMI with multivessel disease, COMPLETE showed fewer cardiovascular death/MI and ischemia-driven events with staged complete revascularization than culprit-only PCI (Mehta 2019, PMID 31475795). Timing may be immediate or staged in stable patients; MULTISTARS AMI informs that decision (Stähli 2023, PMID 37634190). Cardiogenic shock is a separate population in which routine immediate multivessel PCI can add harm.

Procedural outcomes that must be counted

Early outcomes Late outcomes
Death, stroke, MI, major bleeding All-cause and cardiovascular death
Acute kidney injury Spontaneous MI
Vascular complication Repeat revascularization
Periprocedural MI definition Graft occlusion, restenosis, stent thrombosis
Length of stay and delirium Angina, function, quality of life

Periprocedural MI definitions can change composite conclusions and have been central to the left-main controversy. Report each component and time horizon rather than relying on a single composite.

Heart Team and shared decision-making

The Heart Team is most valuable when reasonable strategies diverge: complex left-main/multivessel disease, diabetes, low EF, high surgical risk, frailty, or uncertain completeness. The discussion should provide individualized estimates of early death/stroke/bleeding, long-term MI/reintervention, symptom relief, recovery time, and the consequences of incomplete revascularization (Virani 2023, PMID 37471501; Vrints 2024, PMID 39210710).

Transportability checklist for revascularization trials

Trial feature Bedside question
Run-in medical therapy Was prevention optimized before randomization?
Anatomy excluded Is left main, CTO, diffuse distal disease, or graft disease outside evidence?
Operator/center volume Can local outcomes reproduce trial expertise?
Completeness definition Anatomic, functional, or both?
MI definition Are procedural and spontaneous MI reported separately?
Crossover Strategy outcome or protocol deviation?
Follow-up horizon Has early surgical hazard had time to be offset?
Patient-reported outcomes Was baseline symptom burden sufficient to improve?
Background antithrombotics Does bleeding risk match current regimen?
Surgical candidacy Was PCI compared with CABG in patients truly eligible for both?

Applying this checklist explains why STICHES cannot be used as a direct CABG-versus-PCI estimate and why stable-strategy trials do not settle high-complexity left-main decisions (Velazquez 2016, PMID 27040723; Sabatine 2021, PMID 34793745).

Comparative effect estimates and tradeoffs

Clinical setting Quantified evidence Decision implication
Stable CAD, routine PCI COURAGE death/MI: 19.0% PCI+OMT versus 18.5% OMT (HR 1.05, 95% CI 0.87–1.27) (Boden 2007, PMID 17387127). PCI is not a routine prognostic add-on to good medical therapy.
Moderate/severe ischemia ISCHEMIA 5-year primary-event estimates: 16.4% invasive versus 18.2% conservative (difference −1.8 points, 95% CI −4.7 to 1.0); mortality HR 1.05 (95% CI 0.83–1.32) (Maron 2020, PMID 32227755). Early procedural and later spontaneous MI must be separated.
Diabetes/multivessel disease Individual-patient pooled trials favor CABG survival in diabetes (Head 2018, PMID 29478841); FREEDOM health-status advantages at two years were small and largely attenuated thereafter (Abdallah 2013, PMID 24129463). Prognosis, recovery time, stroke, repeat procedures, and symptoms need separate presentation.
Left main At 5 years in EXCEL the primary composite of death/stroke/MI did not differ (22.0% PCI vs 19.2% CABG; difference 2.8 points, 95% CI −0.9 to 6.5), but all-cause death was higher with PCI (13.0% vs 9.9%; difference 3.1 points, 95% CI 0.2 to 6.1) and ischemia-driven revascularization more frequent (16.9% vs 10.0%) (Stone 2019, PMID 31562798). Pooled individual data across SYNTAX, PRECOMBAT, NOBLE and EXCEL found no significant 5-year mortality difference (11.2% vs 10.2%; HR 1.10, 95% CI 0.91–1.32) with more spontaneous MI (HR 2.35) and repeat revascularization (HR 1.78) after PCI (Sabatine 2021, PMID 34793745). Trial eligibility, SYNTAX complexity, procedural-MI definition, and surgical risk limit generalization.
Ischemic LV dysfunction Long-term STICH supports CABG, with CV-mortality benefit consistent across age despite diminishing all-cause benefit in older patients (Petrie 2016, PMID 27573034). REVIVED found no PCI prognostic benefit, and viability extent did not identify an interaction (Perera 2022, PMID 36027563; Perera 2023, PMID 37878295). CABG and PCI are distinct interventions; “viable myocardium” alone is not a route selector.
Physiology-guided PCI In a study-level reconstruction of the two randomized programs at 5 years, iFR-guided care had higher MACE (HR 1.18, 95% CI 1.04–1.34) and all-cause mortality (HR 1.34, 95% CI 1.08–1.67) than FFR-guided care, with similar MI and unplanned revascularization (Eftekhari 2023, PMID 37634144); ORBITA physiology predicted placebo-controlled symptom response (Foley 2025, PMID 39462291). The late iFR/FFR signal conflicts with earlier noninferiority framing and requires explanation; lesion physiology predicts ischemic potential but does not guarantee symptom or survival benefit.
Multivessel MI without shock COMPLETE reduced CV death/MI with complete revascularization (Mehta 2019, PMID 31475795), and MULTISTARS AMI informs immediate versus staged timing (Stähli 2023, PMID 37634190). The timing result does not apply unchanged to shock or severe renal risk.
Multivessel MI with shock CULPRIT-SHOCK favored culprit-only initial PCI for death/renal replacement (RR 0.83, 95% CI 0.71–0.96) (Thiele 2017, PMID 29083953). Hemodynamic phenotype reverses the intuitive complete-PCI rule.

Why procedural-MI definitions matter

PCI and CABG cause biomarker release with different thresholds and prognostic meaning; spontaneous MI and repeat revascularization emerge later. A composite can therefore favor or disfavor a strategy depending on definitions and follow-up. The ten-year SYNTAX program and left-main joint review explicitly expose how anatomy and endpoint definition change interpretation (Thuijs 2019, PMID 31488373; Byrne 2023, PMID 37632766). Reporting must include all-cause and CV death, spontaneous and procedural MI, stroke, repeat revascularization, renal injury, major bleeding, symptom/function, recovery time, and graft/stent durability.

EXCEL's original randomized comparison established the modern left-main evidence frame before the five-year controversy over its higher all-cause mortality with PCI and its protocol definition of procedural MI (Stone 2016, PMID 27797291; Stone 2019, PMID 31562798). Its eligibility and selected anatomy are as important as the headline result: extrapolation to high-complexity disease or poor procedural expertise is not justified.

The DEFINE-FLAIR/iFR-SWEDEHEART randomized-program meta-analysis provides an earlier mortality/MI comparison for iFR- versus FFR-guided strategies, complementing longer follow-up (Berry 2017, PMID 28972006). Physiology choice does not answer the separate question of PCI versus medical therapy.

Cross-trial synthesis spanning COURAGE, BARI 2D, FAME 2, and ISCHEMIA reinforces medical therapy as the initial strategy for most chronic syndromes while preserving anatomy- and symptom-specific revascularization (Brown 2022, PMID 34767134).

Time horizon separates symptom, repeat-procedure, and survival effects

FAME 2's five-year result illustrates composite interpretation: FFR-guided PCI reduced death/MI/urgent revascularization from 27.0% to 13.9% (HR 0.46, 95% CI 0.34–0.63), driven by urgent revascularization (6.3% vs 21.1%; HR 0.27, 95% CI 0.18–0.41), while death was 5.1% versus 5.2% and MI 8.1% versus 12.0% (HR 0.66, 95% CI 0.43–1.00) (Xaplanteris 2018, PMID 29785878).

The original 1,800-patient SYNTAX comparison found 12-month major cardiac/cerebrovascular events of 17.8% after PCI versus 12.4% after CABG, largely because repeat revascularization was 13.5% versus 5.9%; stroke moved oppositely (0.6% vs 2.2%) (Serruys 2009, PMID 19228612). For selected left-main disease, a four-RCT meta-analysis found no five-year mortality difference (RR 1.03, 95% CI 0.79–1.33), lower stroke with PCI through one year (RR 0.44, 95% CI 0.24–0.82), and higher MI after one year (RR 3.31, 95% CI 2.11–5.18) (Akintoye 2021, PMID 32921597).

ISCHEMIA adds the patient-reported estimand. Invasive care improved the SAQ summary score by 4.1 points (95% credible interval 3.2–5.0) at three months and 2.9 points (2.2–3.7) at 36 months overall; benefit was larger with daily/weekly angina and minimal without baseline angina (Spertus 2020, PMID 32227753). Anatomy, time horizon, and baseline symptoms must therefore remain visible when selecting a route.

Open questions

  • Does any prospectively identifiable stable-disease subgroup gain survival from PCI? (Hochman 2023, PMID 36335918)
  • Can viability, scar burden, physiology, and anatomy identify ischemic-LV patients who benefit from PCI despite neutral REVIVED results? (Perera 2022, PMID 36027563)
  • Will direct CABG-versus-PCI testing in STICH3C resolve route selection in ischemic cardiomyopathy? (NCT05427370)
  • Which endpoint definitions best capture clinically meaningful PCI/CABG trade-offs in left-main trials? (Sabatine 2021, PMID 34793745)
  • Can physiology and imaging guide complete revascularization without increasing unnecessary lesions treated? (Eftekhari 2023, PMID 37634144)

References

  1. Maron DJ, et al. Initial Invasive or Conservative Strategy for Stable Coronary Disease. N Engl J Med. 2020;382:1395-1407. PMID 32227755
  2. Rajkumar CA, et al. A Placebo-Controlled Trial of Percutaneous Coronary Intervention for Stable Angina. N Engl J Med. 2023;389:2319-2330. PMID 38015442
  3. Head SJ, et al. Mortality after coronary artery bypass grafting versus percutaneous coronary intervention with stenting for coronary artery disease: a pooled analysis of individual patient data. Lancet. 2018;391:939-948. PMID 29478841
  4. Stone GW, et al. Everolimus-Eluting Stents or Bypass Surgery for Left Main Coronary Artery Disease. N Engl J Med. 2016;375:2223-2235. PMID 27797291
  5. Sabatine MS, et al. Percutaneous coronary intervention with drug-eluting stents versus coronary artery bypass grafting in left main coronary artery disease: an individual patient data meta-analysis. Lancet. 2021;398:2247-2257. PMID 34793745
  6. Velazquez EJ, et al. Coronary-Artery Bypass Surgery in Patients with Ischemic Cardiomyopathy. N Engl J Med. 2016;374:1511-20. PMID 27040723
  7. Perera D, et al. Percutaneous Revascularization for Ischemic Left Ventricular Dysfunction. N Engl J Med. 2022;387:1351-1360. PMID 36027563
  8. Boden WE, et al. Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med. 2007;356:1503-16. PMID 17387127
  9. Hochman JS, et al. Survival After Invasive or Conservative Management of Stable Coronary Disease. Circulation. 2023;147:8-19. PMID 36335918
  10. Thuijs DJFM, et al. Percutaneous coronary intervention versus coronary artery bypass grafting in patients with three-vessel or left main coronary artery disease: 10-year follow-up of the multicentre randomised controlled SYNTAX trial. Lancet. 2019;394:1325-1334. PMID 31488373
  11. Azzalini L, et al. Percutaneous Coronary Intervention or Surgery for Unprotected Left Main Disease: EXCEL Trial at 5 Years. Interv Cardiol Clin. 2020;9:419-432. PMID 32921366
  12. Byrne RA, et al. 2022 Joint ESC/EACTS review of the 2018 guideline recommendations on the revascularization of left main coronary artery disease in patients at low surgical risk and anatomy suitable for PCI or CABG. Eur J Cardiothorac Surg. 2023;64:ezad286. PMID 37632766
  13. Eftekhari A, et al. Instantaneous wave free ratio vs. fractional flow reserve and 5-year mortality: iFR SWEDEHEART and DEFINE FLAIR. Eur Heart J. 2023;44:4376-4384. PMID 37634144
  14. Berry C, et al. Meta-Analysis of Death and Myocardial Infarction in the DEFINE-FLAIR and iFR-SWEDEHEART Trials. Circulation. 2017;136:2389-2391. PMID 28972006
  15. Foley MJ, et al. Fractional Flow Reserve and Instantaneous Wave-Free Ratio as Predictors of the Placebo-Controlled Response to Percutaneous Coronary Intervention in Stable Coronary Artery Disease. Circulation. 2025;151:202-214. PMID 39462291
  16. Mehta SR, et al. Complete Revascularization with Multivessel PCI for Myocardial Infarction. N Engl J Med. 2019;381:1411-1421. PMID 31475795
  17. Stähli BE, et al. Timing of Complete Revascularization with Multivessel PCI for Myocardial Infarction. N Engl J Med. 2023;389:1368-1379. PMID 37634190
  18. Virani SS, et al. 2023 AHA/ACC/ACCP/ASPC/NLA/PCNA Guideline for the Management of Patients With Chronic Coronary Disease: A Report of the American Heart Association/American College of Cardiology Joint Committee on Clinical Practice Guidelines. Circulation. 2023;148:e9-e119. PMID 37471501
  19. Vrints C, et al. 2024 ESC Guidelines for the management of chronic coronary syndromes. Eur Heart J. 2024;45:3415-3537. PMID 39210710
  20. Abdallah MS, et al. Quality of life after PCI vs CABG among patients with diabetes and multivessel coronary artery disease: a randomized clinical trial. JAMA. 2013;310:1581-90. PMID 24129463
  21. Stone GW, et al. Five-Year Outcomes after PCI or CABG for Left Main Coronary Disease. N Engl J Med. 2019;381:1820-1830. PMID 31562798
  22. Petrie MC, et al. Ten-Year Outcomes After Coronary Artery Bypass Grafting According to Age in Patients With Heart Failure and Left Ventricular Systolic Dysfunction: An Analysis of the Extended Follow-Up of the STICH Trial (Surgical Treatment for Ischemic Heart Failure). Circulation. 2016;134:1314-1324. PMID 27573034
  23. Perera D, et al. Viability and Outcomes With Revascularization or Medical Therapy in Ischemic Ventricular Dysfunction: A Prespecified Secondary Analysis of the REVIVED-BCIS2 Trial. JAMA Cardiol. 2023;8:1154-1161. PMID 37878295
  24. Thiele H, et al. PCI Strategies in Patients with Acute Myocardial Infarction and Cardiogenic Shock. N Engl J Med. 2017;377:2419-2432. PMID 29083953
  25. Brown DL, et al. Optimal Medical Therapy as First-Line Therapy for Chronic Coronary Syndromes: Lessons from COURAGE, BARI 2D, FAME 2, and ISCHEMIA. Cardiovasc Drugs Ther. 2022;36:1039-1045. PMID 34767134
  26. Serruys PW, et al. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med. 2009;360:961-72. PMID 19228612
  27. Xaplanteris P, et al. Five-Year Outcomes with PCI Guided by Fractional Flow Reserve. N Engl J Med. 2018;379:250-259. PMID 29785878
  28. Akintoye E, et al. Intermediate and Late Outcomes With PCI vs CABG for Left Main Disease - Landmark Meta-Analysis of Randomized Trials. Cardiovasc Revasc Med. 2021;23:114-118. PMID 32921597
  29. Spertus JA, et al. Health-Status Outcomes with Invasive or Conservative Care in Coronary Disease. N Engl J Med. 2020;382:1408-1419. PMID 32227753