Better Survival for CABG vs PCI in Some High-Risk Patients: ASCERT

Reed Miller

January 30, 2012

January 30, 2012 (Fort Lauderdale, Florida)ASCERT, a large comparative-effectiveness study derived from Medicare and professional society databases, found that CABG surgery provides better four-year survival odds than PCI in high-risk stable patients with two- or three-vessel disease [1].

Dr Fred Edwards Source: ASCERT

"We still don't have the answer to the question of how patients with stable coronary artery disease should be treated if they don't respond to medical therapy," ASCERT primary investigator Dr Fred Edwards (University of Florida, Jacksonville) told heartwire .So the Society of Thoracic Surgeons (STS),the American College of Cardiology (ACC), and the Centers for Medicare & Medicaid Services (CMS) collaborated to develop ASCERT, a comparative-effectiveness study of surgery vs PCI in stable coronary atherosclerosis patients.

The ASCERT high-risk subset results, presented here at the Society of Thoracic Surgeons 2012 Annual Meeting, show a long-term survival benefit for surgery over percutaneous intervention. "This confirms, in current real-world practice, the results of other studies, from the New York state data to the randomized trials like SYNTAX," STS president Dr Michael Mack (Baylor Health Care System, Plano, TX) told heartwire . "That's what we're finding in the real world under current medical therapy. This is encouraging information, in that the results of previous trials are still relevant to current populations.

Dr Michael Mack

"The whole reason that the ACC and STS formed their databases to begin with is that administrative data are not able to adjust for risk . . . but one of the shortcomings of that was that we could capture only 30-day outcomes and couldn't get long-term follow-up," Mack said. "So linking this to CMS data is huge in terms of all of this research long term. It has the best of both worlds. It's risk-adjusted and has long-term outcomes."

The study combines PCI data from the National Cardiovascular Data Registry (NCDR), bypass surgery data from the STS database, and up to four years of Medicare outcomes data (average 2.67 years) from CMS. Overall, the study includes 86 244 bypass-surgery patients and 103 549 PCI patients treated from 2004 through 2007. The mean age was 74 years. The high-risk subset represented about 2% of the patients in the trial, Edwards said.

Survival rates favored percutaneous intervention within one year--about 1% vs 2% mortality for percutaneous intervention vs surgery, respectively. But after one year, bypass surgery was associated with progressively better survival than percutaneous intervention. For high risk patients--75 or older, diabetic, ejection fraction <50%, and glomerular filtration rate <60 mL/min/1.73m2--bypass surgery was associated with lower four-year mortality than PCI (risk ratio=0.72).

"The results should improve the quality of care for patients with coronary disease and should clarify the indications for the subgroups [analyzed in the study]," Edwards said.

The complete ASCERT results will be presented at the American College of Cardiology 2012 Scientific Sessions in Chicago in March.

The authors also conducted multiple propensity analyses to root out any selection bias or unmeasured confounders. The statistical analysis found that if there is an unknown confounder, it would have to have an effect as big as diabetes or end-stage renal failure to produce the difference between survival curves seen in this trial, Edwards said.

"Now the key is for referring physicians, cardiologists, and surgeons to avail themselves of this new information about long-term survival and make sure that that's discussed with the patients and all of the care providers of the patients to decide what the optimal course of therapy is," Edwards said.

Is This a Comparison or Just a Description?

Dr Christopher White

President of the Society for Cardiovascular Angiography and Interventions (SCAI), Dr Christopher White (Ochsner Medical Center, New Orleans, LA), told heartwire that although SCAI supports ASCERT, the study should not be interpreted as a direct comparison between the two therapies.

"This is not a comparison. This was never comparison of patients. This is a description in a database of what happens when patients are treated one way or another," he said. "What this report really says is that we are really good at picking out which patients do better with surgery and which patients do better with stenting. They are not the same patients getting those treatments. So it's crazy for someone to say that surgery is better than stenting. That's not what this is about. When see a patient in my clinic and I think they have problems that warrant surgery, I refer them there and they get surgery, but when I see someone who needs stenting, I do that," he said.

"There's a whole number of patients who are too ill and too high risk to get surgery, so those patients are offered [PCI] and often the reasons for not doing surgery is that the patients have comorbidities . . . which also kill them," he said. "There's no way to risk-adjust that kind of information because the outcomes are too divergent.

"If you really want to know whether surgery is better than stenting, you have to do the randomized trial, and those are out there. That's what SYNTAX is, [for example]," he said.

The First Fruit of an Important Collaboration

ASCERT represents over 10 times as many patients as the total enrollment of the randomized trials comparing bypass surgery and PCI in this population, and this is only now available because the STS, ACC, and CMS were able to work together to pool their data. The study is one of the comparative-effectiveness studies sponsored by the National Institutes of Health through a "grand opportunity" created by the American Recovery and Reinvestment Act of 2009, Edwards said.

Mack added: "The fact that there are two databases married to CMS data is hugely [important]. This is the first toe in the water of being able to do honest-to-goodness robust comparative-effectiveness research, and it's not only just this first information that comes out of it but what it is going to lead to down the line." Mack pointed out that the TVT Registry of transcatheter valves is a "son or daughter of ASCERT" because it evolved out of this ACC/STS/CMS collaboration. Edwards is also leading that registry.

In addition to the long-term survival comparison, Edwards's group plans to look at stroke rates, hospital-readmission rates, and cost effectiveness to create efficiency measures. "Survival is only one part of the picture," Edwards said during the presentation of these results at the STS meeting. "We shouldn't say 'Well, you got better survival with this intervention or the other, and therefore everybody should have the one with better survival.' Quality of life clearly needs to be taken into account."

They are also going to compare coronary bypass surgery and percutaneous intervention SYNTAX score strata and the therapies in patients with single-vessel disease. But White said he doubts that analysis will reveal anything beyond confirming the original SYNTAX results. "The patients who get stents are more likely to die regardless of what you do to them within two years than the people you give surgery. I'm not sure anything will be able to balance that as a risk adjustment," he said. "The SYNTAX trial was already done in randomized people where you really do get a comparison."

Edwards said they're also considering conducting a similar study in patients with left main disease, but the small numbers and selectivity of those patients may preclude that study.

Edwards is a consultant for Humana. Mack has a nonremunerative position of interest on the executive committee of the PARTNER trial for Edwards Lifesciences. White is on the advisory boards of St Jude, Neovasc, and Baxter Cellular Therapy.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.