Gregg W. Stone (Icahn School of Medicine, New York, USA)clarified how to translate the final EXCEL trial outcome into real-world recommendation for patients.
As the largest left main intervention trial to date, the EXCEL trial recruited 1,905 patients with unprotected left main disease with a SYNTAX score <32, and upon evaluation, were randomized to either PCI with Xience EES or CABG.
The primary endpoint was death, MI or stroke at a median of three years. Results showed no significant difference in the primary endpoint between PCI (15.4%) and CABG (14.7%).
Stone stressed that this was the only endpoint that EXCEL was powered to show and other findings such as the five-year endpoint were underpowered or exploratory. The five-year endpoint showed no significant difference (p=0.13) between PCI and CABG but the curves crossed over, indicating PCI had less events earlier in the trial but more events as time progressed.
There also appeared to be three distinct periods of varying relative risk: 0-day to 30-days where PCI was better than CABG by about 40 percent; 30-days to one-year where they were the same; and one-year to five-years where CABG was better than PCI.
Overall, at the end of five-year follow up, the event free survival time was 5.2 days longer after PCI compared to CABG. Clearly, there was no major difference in death, MI or stroke between PCI and CABG in the five-year follow up.
“We can translate the EXCEL findings to the patient via a heart team that utilizes the available data and explain the risk and benefits of the procedures to the patient,” Stone said. “The patient can then make an informed decision based on their own preference.”
Interestingly, patients and physicians had different perspectives when rating the importance of clinical outcomes. Most physicians rated death as the most serious adverse outcome while stroke and MI as less important. Patients, however, rated stroke and MI equally serious or even more serious than death. Overall, death, stroke and MI were mutually important factors in the shared decision-making process.
Stone then presented the death rates, stroke rates, myocardial infarction (MI) rates and quality of life measures in the PCI group compared to CABG group.
In terms of mortality rates, there was no difference between PCI and CABG at five- and 10- year follow-up. Because no single trial was powered for five-year mortality in LM PCI vs CABG, meta-analysis of NOBLE, SYNTAX, PRECOMBAT and EXCEL trials were used.
Meta-analysis showed no significant difference between PCI and CABG. Although CABG is often perceived as better than PCI with respect to long-term mortality benefit, this was not observed in the trials.
The 10-year mortality data in the SYNTAX left main trial for CABG was 26.7 percent and 26.1 percent for PCI, indicating no significant difference at 10 years. The PRECOMBAT trial also showed no statistical difference in 10-year mortality (14.5% PCI vs. 13.8% CABG)
The relative risk of stroke was lower by two third with PCI compared to CABG. However, the absolute difference was small at 0.8 percent at one year and not statistically significant at five years. Therefore, stroke rates should not be the only reason in treatment decision-making unless they are at particularly high risk of stroke, Stone said.
There was also no overall difference in the rate of myocardial infarction between PCI and CABG. However, there was a difference in the timings of the infarctions with fewer procedural MIs after PCI and fewer non-procedural MIs or late-MI after CABG.
“CABG is a durable procedure and bypass grafting does treat more disease and address vulnerable plaques better than PCI,” Stone said.
Quality of life was better with PCI as evidenced by the 72 percent reduction of major adverse events within 30-days in PCI compared to CABG in the EXCEL trial. These events included death, MI, bleeding, transfusions, major arrhythmias, cardioversions, pacemakers, repeat procedures, renal failure, sepsis, and prolonged intubation.
In terms of physical summary scale outcomes, patients post-PCI felt significantly better than post-CABG patients at one month. However, no difference between PCI and CABG was observed after one and three years. Both PCI and CABG were highly effective in relieving angina.
Stone then outlined the advantages and drawbacks for each procedure.
In PCI, there are early advantages of the procedure being less invasive with fewer peri-procedural complications (stroke, MI, atrial fibrillation, bleeding, acute kidney injury, etc.), lower 30-day MACE and more rapid recovery with better quality of life and earlier angina relief.
CABG has late advantages, with the procedure being more durable and fewer adverse events beyond one year, particularly MI and repeat revascularization procedures. In the majority of patients, however, there will not be a major difference in long term survival, MACE (death, MI or stoke) and quality of life.
Stone concluded by saying: “At the end of the day, the choice between PCI and CABG depends on the state of the coronary vessels, patient’s preference and the doctor’s experience and expertise. There are some who clearly benefit from PCI rather than CABG and vice versa. However, many patients fall in between the two. Heart team discussions are important and the patients should be aware that PCI and CABG are two very different procedures. Everything should be communicated to the patient and his/her family in making the final decision.”
Editor: Benjamin Leo Cheang Leng,MD (Columbia Asia Hospital Tebrau, Malaysia)