December 13th, 2012

Diabetics with Multivessel Disease: FREEDOM with CABG?

FREEDOM (Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease), a large NIH-sponsored study presented at AHA 2012 and published simultaneously in the New England Journal of Medicine, showed that diabetics with multivessel disease had lower rates of death and MI with CABG than with PCI, but a higher rate of stroke. Study author Dr. Valentin Fuster answered CardioExchange’s questions about FREEDOM and what it means for patients.

Click here to listen to the original audio interview.

Q: What will you tell your patients with diabetes who would have met inclusion criteria for FREEDOM? How will you help them to understand their options?

A: The problem is that most patients with diabetes who undergo cardiac catheterization already have two- or three-vessel disease. We have to tell the patients prior to the catheterization that the angiogram will probably show disease for which surgical intervention is the option of choice. This is a different approach from what we’ve been doing, which is when catheterization is often immediately followed by a decision to do an angioplasty or insert a stent.  Patients need to be informed of their options beforehand.

Q: So the point here is to prevent ad hoc PCI?

A: Yes, it’s beyond whether a patient is a candidate or not. It usually begins in the cath lab. The main reason that people were rejected from our study was because they indicated at the time of the procedure that they would prefer an angioplasty over surgery, which ruled them out as a participant.

Q: What was the nature of the strokes in FREEDOM?

A: Of the patients who had them, 60% of those in the PCI group and 70% of those in the CABG group experienced some degree of disability , and 27% of those in the PCI group and 55% of those in the CABG group experienced a significant degree of disability. The number of strokes was relatively small and they occurred throughout the study, rather than only at  the time of the procedure. We actually believe that there were fewer strokes in the PCI group over the long term because they were taking clopidogrel and aspirin together, whereas the majority of the CABG group were taking aspirin alone. We are planning to test this hypothesis.

Q: Was there any heterogeneity in outcome by the diabetes severity?

A: Most of the patients had type-2 diabetes, and the average hemoglobin A1c level was 7.8%, however, there was a large group that had more than 7% and another group that had less than 7%. Regardless of the degree of diabetes control, the results were equal in favor of CABG.

Q: Do you think the results are relevant to patients in their 80s — for whom CABG might be a bigger deal?

A: There were very few patients who were age 80 or older. I would say that at that advanced age the concern of survivorship, myocardial infarction, and stroke are secondary compared with whether or not you can relieve the patient’s angina. For a patient at that age, I would perhaps give a second thought to a stent for specific individuals after full discussion. I’m not necessarily in favor of stenting, but, depending on what your goal is, it should be discussed. If the goal is to relieve angina that might be accomplished with simple stenting.

2 Responses to “Diabetics with Multivessel Disease: FREEDOM with CABG?”

  1. David Powell , MD, FACC says:

    The robust findings will change practice. Nevertheless, I have two points:

    The substantial group excluded after angiography should be followed carefully and analyzed. These patients “preferred stenting”. But decisions are made with the angiographer. We should know that this group was similar to the randomized group, not skewed toward lower SYNTAX scores, for example. I understand that the SYNTAX score did not interact with the surgical benefit among the randomized group. Also, one might expect that the excluded group would fare poorly with respect to MI and survival.

    There was no medical arm, and some of these patients satisfied entry criteria for COURAGE. Yes, BARI 2D demonstrated a benefit for revascularization over medical therapy in those diabetics whose anatomy was most amenable for CABG. One may argue that those BARI 2 patients who got PCI after randomization to the revascularization arm would have fared better with CABG than medical therapy. But this is somewhat speculative. A subset of FREEDOM patients may fare better with medical therapy, particularly those at higher stroke risk.

    Lastly, what are the implications for ongoing trials like ISCHEMIA? Should the diabetics with 3vd be excluded from this PCI study and get CABG?

  2. Dr. Powell raises some great questions. In response, I’m writing to address the design and rationale of ISCHEMIA and place it in the context of BARI 2D and FREEDOM. First, ISCHEMIA is not strictly a “PCI study.” ISCHEMIA will compare the effectiveness of two initial management strategies – conservative (optimal medical therapy alone with selective cardiac catheterization and revascularization for those who fail medical management) vs. invasive (optimal medical therapy plus cardiac catheterization followed by optimal revascularization with PCI or CABG, whichever is most appropriate) – in 8,000 patients with stable ischemic heart disease (SIHD) and at least moderate ischemia. The decision regarding the method of revascularization will rely heavily on the judgment of the Heart Team at each enrolling site. In light of the FREEDOM trial results, CABG will be strongly favored as the method of revascularization in participants with diabetes randomized to the invasive strategy who are found to have 3-vessel disease.

    Second, as Dr. Powell points out, there was no medical arm in FREEDOM, but there was in BARI 2D. BARI 2D found that an initial management strategy of revascularization plus medical therapy (MT) did not reduce the overall rate of the primary endpoint (death) as compared with MT alone. Survival was not improved in either the CABG or PCI subsets. The principal secondary endpoint was the composite of death, MI, and stroke. There was no overall difference in this secondary endpoint between revascularization and medical therapy, but it was lower (due to a lower MI rate) in the CABG stratum assigned to CABG (n=378) vs. those assigned to MT (n=385, 5-year event rate 22.4% vs. 30.5% [79 vs. 115 events], p=0.01). Only 230 patients were followed out to 5 years. In the context of an overall “negative” trial, all subgroup data must be viewed as hypothesis generating. The findings in the CABG vs. MT subset analysis of the secondary endpoint is interesting but not robust enough to be definitive and guide routine clinical practice. Substantially more robust data are needed and will be provided by the ISCHEMIA trial.

    As for FREEDOM, without a direct randomized comparison of CABG with MT one can only speculate whether CABG is superior to contemporary comprehensive secondary prevention. Furthermore, unlike ISCHEMIA, all FREEDOM patients underwent cath prior to randomization. It should be noted that 30% of FREEDOM patients had recent ACS prior to enrollment, whereas recent ACS is an exclusion criterion for ISCHEMIA. To reduce the likelihood of cardiac catheterization and revascularization in participants randomized to the conservative strategy, ISCHEMIA will also exclude patients who have an unacceptable level of angina despite maximal medical therapy or who are very dissatisfied with their medical management of angina.

    So although there is evidence to suggest that patients with diabetes and 3-vessel disease may have better outcomes with CABG as compared with medical therapy, this has not been demonstrated in the context of a randomized trial using contemporary medical therapy, and this argues for the inclusion of SIHD patients with diabetes in the ISCHEMIA trial.

    A final point: Based on current evidence (summarized above) it is possible that CABG reduces the risk of MI in patients with diabetes and multivessel disease. Therefore a fundamental consideration for this patient population when reviewing treatment options with their physicians is whether it is worth taking the small immediate risks of perioperative death and stroke for the potential reduction of future MI events without evidence of a reduction in long-term mortality.