Pre-reading

STUDY DESIGN - Patients were randomized in a 1:1 fashion to either PCI with Xience everolimus-eluting stents (n = 948) or CABG (n = 957).

Total number screened: 2,905

Total number of enrollees: 1,905

Duration of follow-up: 3 years

Mean patient age: 66 years

Percentage female: 23%

INCLUSION CRITERIA

- Unprotected left main coronary artery disease with ≥70% stenosis, or ≥50% to <70% with either: 1) noninvasive evidence of left main ischemia, 2) intravascular ultrasound minimum lumen area ≤6.0 mm2, or 3) fractional flow reserve ≤0.80

- SYNTAX score ≤32

- Clinical and anatomic eligibility for both PCI and CABG, as agreed to by the local Heart Team

EXCLUSION CRITERIA

Prior CABG or left main PCI any time

Prior non left main PCI within 1 year

Need for cardiac surgery other than CABG

Inability to tolerate dual antiplatelet therapy (DAPT) for 1 year

Creatine kinase-myocardial band > upper limit of normal

OTHER SALIENT FEATURES/CHARACTERISTICS

Diabetics: 29%

Prior PCI: 17%

Stable angina, 53%; recent myocardial infarction (MI), 15%; unstable angina, 24%

Mean left ventricular ejection fraction (LVEF): 57%

SYNTAX scores: low (≤22) for PCI vs. CABG, 32.2% vs. 39.3%; high (≥33), 25.1% vs. 23.4%

Distal left main trunk bifurcation lesion: 81%

Additional single-vessel disease, 31%; two-vessel disease, 33%; three-vessel disease, 18%

PCI arm: IVUS, 77.2%; hemodynamic support, 5.2%; stents per patient, 2.4; stented length, 49.1 mm

CABG arm: off-pump, 29.4%; conduits per patient, 2.6; internal mammary artery use, 99%

Medication use: aspirin, 99%; statin, 95%; P2Y12 inhibitor for PCI vs. CABG, 97.6% vs. 32.6%; p < 0.001

PRINCIPAL FINDINGs

The primary outcome, death, stroke, or MI at 3 years for PCI vs. CABG was 15.4% vs. 14.7%,  p = 0.018 for noninferiority, p = 0.98 for superiority.

Death: 8.2% vs. 5.9%, p = 0.11

Stroke: 2.3% vs. 2.9%, p = 0.37

MI: 8.0% vs. 8.3%, p = 0.64

SECONDARY OUTCOMES

Death, stroke, or MI at 30 days for PCI vs. CABG: 4.9% vs. 7.9%, p = 0.008

Death, stroke, or MI between 30 days and 3 years for PCI vs. CABG: 11.5% vs. 7.9%, p = 0.02

Stent thrombosis or graft occlusion at 30 days: 0.3% vs. 1.2%, p = 0.03

Ischemia-driven revascularization at 3 years: 12.6% vs. 7.5%, p < 0.0001

Stent thrombosis or graft occlusion at 3 years: 0.7% vs. 5.4%, p < 0.001

QUALITY-OF-LIFE ASSESSMENTS - There was no difference in the Seattle Angina Questionnaire frequency between the PCI and CABG arms at 12 months (difference = -0.3, p = 0.63) or 36 months (difference = -0.8, p = 0.21). No differences were noted based on baseline SYNTAX scores at 1 year, but low SYNTAX score patients seemed to do better with CABG at 3 years (p for interaction = 0.03). More patients scored 0 (none) on the Rose Dyspnea Scale at 1 month with PCI than CABG (60% vs. 44%, p < 0.01), but this difference was not significant on subsequent follow-up. Similarly, the Short Form-12 physical summary scale was significantly higher for PCI at 1 month (difference = 8.2, p < 0.01), but similar at 36 months (difference = -0.4, p = 0.5). Clinically significant depression was also lower with PCI at 1 month (8% vs. 19%, p < 0.01) and 12 months (8% vs. 12%, p = 0.03), but similar at 36 months (9% vs. 8%, p = 0.77).

NEW-ONSET ATRIAL FIBRILLATION (NOAF) - Overall incidence was 6.7%, almost entirely among patients undergoing CABG (161 out of 162 patients with NOAF; incidence post-CABG: 18%). By the time of hospital discharge, NOAF had resolved in 85.8%. Overall, 10.1% were discharged on an oral anticoagulant. DAPT use in the CABG arm was 33% on discharge. By multivariable analysis, NOAF after CABG was an independent predictor of 3-year stroke (6.6% vs. 2.4%, adjusted hazard ratio [HR] 4.19, 95% confidence interval [CI] 1.74-10.11, p = 0.001), death (11.4% vs. 4.3%, adjusted HR 3.02, 95% CI 1.60-5.70, p = 0.0006), and the primary composite endpoint of death, MI, or stroke (22.6% vs. 12.8%, adjusted HR 2.13, 95% CI 1.39-3.25, p = 0.0004).

ROLE OF B-TYPE NATRIURETIC PEPTIDE (BNP) - BNP at baseline was elevated (≥100 pg/ml) in 39.5% of enrolled patients. Patients with elevated BNP levels were older and more frequently had additional cardiovascular risk factors and lower LVEF than those with normal BNP, but had similar SYNTAX scores. A significant interaction (p interaction = 0.03) was present between elevated versus normal BNP and treatment with PCI versus CABG for the adjusted risk of the primary composite endpoint at 3 years among patients with elevated BNP (21.1% vs. 16.1%, adjusted HR for PCI vs. CABG 1.54, 95% CI 0.96-2.47) versus normal BNP (10.3% vs. 13.1%, adjusted HR 0.74, 95% CI 0.46-1.20).

CHRONIC KIDNEY DISEASE (CKD) PATIENTS - CKD was present in 19.3% of the patient population. Patients with CKD had higher 3-year rates of the primary endpoint compared with those without CKD (20.8% vs. 13.5%, HR 1.60, 95% CI 1.22-2.09, p = 0.0005). Acute renal failure occurred less commonly after revascularization with PCI compared with CABG, both in patients with CKD (2.3% vs. 7.7%, HR 0.28, 95% CI 0.09-0.87) and in those without CKD (0.3% vs. 1.3%, HR 0.20, 95% CI 0.04-0.90, p for interaction = 0.71). There were no significant differences in the rates of the 3-year primary composite endpoint after PCI and CABG in patients with CKD (23.4% vs. 18.1%, HR 1.25, 95% CI 0.79-1.98) and without CKD (13.4% vs. 13.5%, HR 0.97, 95% CI 0.73-1.27, p for interaction = 0.38). Sternal wound dehiscence was high among CKD patients (0% vs. 3.3% for PCI vs. CABG, p = 0.03).

FIVE-YEAR RESULTS - The primary endpoint for PCI vs. CABG was 22.0% vs. 19.2%; odds ratio 1.19 (95% CI 0.95-1.50, p = 0.13). On landmark analysis at 1 year, rates were significantly higher with PCI vs. CABG (15.1% vs. 9.7%, p < 0.001, p for time interaction < 0.001). Higher rates at 5 years with PCI were driven by all-cause mortality (13% vs. 9.9%), and nonprocedural myocardial infarction (6.8% vs. 3.5%). Ischemia-driven target lesion revascularization rates were also higher (16.9% vs. 10%). Stent thrombosis vs. graft occlusion was 1.6% vs. 6.5%. Rates for the primary endpoint were independent of SYNTAX score.

INTERPRETATION - The results of this important trial indicate that PCI with second-generation DES (Xience) is noninferior to CABG for clinical and functional outcomes at 3 years following revascularization of unprotected left main lesions. As noted earlier, repeat revascularization rates were higher with PCI. Thrombosis (stent vs. graft) rates were lower with PCI than with CABG. As expected, adverse clinical events were not uniformly distributed from a temporal standpoint between the two arms. The hazard was highest with CABG in the first 30 days and clinical outcomes were actually better with PCI up to 30 days. However, this reversed between 30 days and 3 years, such that outcomes were inferior with PCI compared with CABG beyond this time frame. This was also noted out to 5 years. Further longer-term follow-up is awaited.

REFERENCES

1. Stone GW, Kappetein AP, Sabik JF, et al., on behalf of the EXCEL Trial Investigators. Five-Year Outcomes After PCI or CABG for Left Main Coronary Disease. N Engl J Med 2019;381:1820-30.

2. Giustino G, Mehran R, Serruys PW, et al. Left Main Revascularization With PCI or CABG in Patients With Chronic Kidney Disease: EXCEL Trial. J Am Coll Cardiol 2018;72:754-65.

3. Editorial Comment: O’Gara PT. PCI or CABG for LMCA Revascularization in Patients With CKD: The Jury Is Still Out. J Am Coll Cardiol 2018;72:766-8.

4. Redfors B, Chen S, Crowley A, et al. B-Type Natriuretic Peptide Assessment in Patients Undergoing Revascularization for Left Main Coronary Artery Disease: Analysis From the EXCEL Trial. Circulation 2018;Apr 17:[Epub ahead of print].

5. Kosmidou I, Chen S, Kappetein AP, et al. New-Onset Atrial Fibrillation After PCI or CABG for Left Main Disease: The EXCEL Trial. J Am Coll Cardiol 2018;71:739-48.

6. Editorial Comment: Verma A, Bhatt DL, Verma S. Long-Term Outcomes of Post-Operative Atrial Fibrillation: Guilty as Charged. J Am Coll Cardiol 2018;71:749-51.

7. Baron SJ, Chinnakondepalli K, Magnuson EA, et al., on behalf of the EXCEL Investigators. Quality of Life After Everolimus-Eluting Stents or Bypass Surgery for Treatment of Left Main Disease. J Am Coll Cardiol 2017;Oct 30:[Epub ahead of print].

8. Editorial Comment: Mark DB, Patel MR. Patient-Reported Outcomes in Revascularization Decisions for Left Main Disease: Sharing the EXCELlence. J Am Coll Cardiol 2017;Oct 30:[Epub ahead of print].

9. Stone GW, Sabik JF, Serruys PW, et al., on behalf of the EXCEL Trial Investigators. Everolimus-Eluting Stents or Bypass Surgery for Left Main Coronary Artery Disease. N Engl J Med 2016;375:2223-5.

10. Editorial: Braunwald E. Treatment of Left Main Coronary Artery Disease. N Engl J Med 2016;375:2284-5.