CABG vs. PCI in low-risk multi-vessel disease
In 1994, Yusuf et al. assessed the effect of CABG on survival by performing a meta-analysis of the results of seven RCT of CABG vs. medical therapy (2650 patients followed for 10 years); results showed that CABG improved survival and symptom relief especially in three-vessel disease (3VD) . Benefits were greater in the presence of severe symptoms, positive exercise ECG and impaired left ventricular function. However, there was no survival benefit for CABG over standard medical therapy if there were single-vessel disease (1VD) or two-vessel disease (2VD) and normal LV function. These investigators recommended for future trials of PCI and GABG to include a high proportion of patients for whom surgery is known to be superior to medical therapy. The crucial question is whether we followed this recommendation or not.
Table 1 shows the summary of 15 RCT of PCI vs. CABG in multivessel disease. It demonstrates that only 5% of screened patients were actually randomized to the two therapies. Importantly, all patients studied have ejection fractions of 50%. The incidence of 3VD in these trials is only 35%. It should not be surprising to find that CABG and PCI are similar in outcomes as these trials exclude all the high-risk patients which may benefit from CABG over multivessel intervention (i.e. diabetics, low LV function and 3VD or left main). However, one might argue these trials reflect the natural proportion of multivessel disease in the population, and an appropriate ratio to be representative of ‘all comers’ with CAD; but then we should be very cautious in generalizing these results to all CAD patients.
|Table 1 Summary of 15 RCT PCI vs. CABG in multivessel disease|
|Trial||No. Screened||% Randomized||Stent||% 3-vessel disease||Proximal LAD||EF>50%||% Diabetes|
RCT, randomized clinical trials; PCI, percutaneous coronary interventions; CABG, coronary artery bypass surgery; LAD, left anterior descending artery; EF, ejection fraction.
Hoffman et al. reported a meta-analysis of 13 RCT CABG vs. PCI . Two-thirds of the patients had 2VD and all patients had normal LV function. High-risk patients i.e. diabetics, decreased LV function and 3VD were not enrolled. Despite being low-risk CAD, CABG resulted in a small survival advantage and a marked reduction in the need for repeat revascularization. There was a small, 1.9%, absolute survival advantage favoring CABG over PCI for all trials at 5 years. However, it must be noted that all data at 5 years are from earlier studies that did not employ stents in the initial revascularization procedure. They also performed subgroup analyses of trials with and without stents in the initial PTCA arm where data were available from at least two trials. This trend favoring CABG disappeared when it was compared to more recent trials with stents. Whereas the risk difference of CABG vs. balloon angioplasty (POBA) repeat revascularization was 34% at 3 years, this difference decreased to 15% when coronary stents were used (2).
To study this further, Mercado et al. published the results of a meta-analysis on CABG vs. PCI (with stents) for multivessel disease in 2005 . Investigators included four RCT (Arterial Revascularization Therapies Study, Stent or Surgery Trial, Argentine Randomized Trial of Percutaneous Transluminal Coronary Angioplasty vs. CABG in Multivessel Disease 2, and Medicine, Angioplasty, or Surgery Study 2) that compared PCI with multiple-vessel stenting (n=1518) vs. CABG (n=1533). Again only 4% of the screened patients were randomized. One year after the initial procedure, PCI with multiple stenting and CABG provided a similar degree of protection against death, myocardial infarction, or stroke in patients with MVD. Repeat revascularization procedures occurred more frequently in patients in the PCI group compared to CABG (18% vs. 4.4%; hazard ratio 4.4 and 95% confidence interval 3.3–5.9) .
Finally, analysis of the ARTS randomized trial revealed the five-year outcomes after PCI with multi-vessel stenting vs. CABG for the treatment of MVD . Only 10% of screened patients were randomized. Again 70% had 2VD and all of them had normal LV function. Even though results of ARTS were already predicted by patient population, one-year and five-year mortality rates were similar in both arms. As expected, repeat revascularization rates were significantly higher in the stent arm. Subgroup analysis showed that patients with diabetes had a survival benefit if they got randomized to CABG. Interestingly, trials excluded patients known to benefit from CABG and patients were not representative of most CABG patients. Nevertheless, trial results have been generalized to all patients.
It is important to note that all the above meta-analyses (2–4) although reported to study patients with multivessel disease, has the majority of patients with 2VD and normal LV function. The results of these meta-analyses replicate the results of initial meta-analysis by Yusuf et al.  published a decade before these trials were undertaken.
In summary, it would be safe to conclude that CABG and PCI are both reasonable options in patients with single or 2VD CAD with normal LV function. There may be a slight mortality benefit long-term of CABG over balloon angioplasty (POBA) which is nullified with the use of stents. However, there is approximately four times more repeat procedures in patients treated initially with PCI compared to CABG which is about half of what is seen with balloon angioplasty .