October 3, 2011 (Lisbon, Portugal) — New data from a large UK observational study have somewhat controversially shown that off-pump coronary artery bypass surgery (CABG) is associated with significantly higher in-hospital and mid-term mortality than conventional on-pump CABG. Dr Domenico Pagano (University Hospital and University of Birmingham, UK) reported the findings in a last-minute addition to the program at the European Association for Cardio-Thoracic Surgery 2011 Annual Meeting here yesterday.
Off-pump CABG "was associated with a 34% increased risk of in-hospital mortality overall and a 28% reduced risk of mid-term survival, out to almost three years, compared with on-pump," Pagano noted. These new data are in line with findings from smaller, randomized studies, says Pagano, adding, "It's a good snapshot of contemporaneous practice in the UK."
But others criticized the study. Dr David Taggart (University of Oxford, UK), said, "The fundamental flaw in what is potentially a dangerous interpretation is that this was done at an institutional level, and that covers many sins, including surgeons who had attempted off-pump surgery--it didn't work out for them, and they abandoned it. I would want to see these data reanalyzed looking at what happened in the outcomes of surgeons who are high-volume users of off-pump surgery. We need propensity-matching of the operators. That's my concern in the interpretation of these data."
Other attendees pointed out further drawbacks, including the fact that off-pump CABG (also known as OP-CAB) is a whole family of approaches, and the way it was performed in this study was not properly defined. In addition, some of the findings could have been driven by graft failure, and there was no allowance for this in the study design.
Asked to comment on the results, Dr John Pepper (Imperial College London, UK) explained to heartwire there have been "many subtle improvements in heart-lung machines" since off-pump CABG first emerged, and now "the pump is pretty safe. Plus you have to remember that many CABG operations are done in conjunction with valve procedures, which are done on a heart-lung bypass machine."
With this background, "My opinion is that off-pump and on-pump are similar," says Pepper. "Off-pump is not significantly better than on-pump, but in experienced hands, in experienced units, they can achieve extremely good results with off-pump. I think the message is that off-pump is only for a group of people who are very keen on doing it, who have trained themselves obsessively, and then they can get results that are comparable to, but probably not much better than, on-pump."
Pepper added that the results of an ongoing large Canadian randomized clinical trial of off-pump vs on-pump surgery in around 4000 patients are eagerly awaited to try to discern whether there truly is a difference between these two approaches in terms of outcomes.
Compared With on-Pump Surgery, 108 Extra Deaths on off-Pump Group
When off-pump CABG first emerged as a surgical technique, it was thought it would reduce or eliminate the neuropsychological risks of putting a patient on a heart-lung bypass machine, avoiding or reducing so-called "pump head." Pagano explained that off-pump CABG may improve outcomes of selected groups of patients, and "there are a number of small randomized trials that show superiority in some subgroups." But "landmark randomized trials to address short-term and long-term mortality are still not available," he noted.
In the new analysis, he and his colleagues examined data on patients undergoing isolated, first-time CABG between March 2004 and April 2009, obtained from the national database for cardiothoracic surgery of Great Britain and Ireland. Mortality data were extracted from the UK Office for National Statistics. While observational data are no substitute for randomized controlled trials, "they still have a role to provide the best available evidence for practice to date," Pagano observed.
The overall patient population consisted of 58 072 patients, of whom 14% (n=9590) underwent off-pump CABG and 86% (n=48 482) had on-pump surgery. Of these, 6374 off-pump patients were propensity-matched with 6374 on-pump patients for the primary analysis. To further reinforce the main analysis, the propensity score was also used as a covariant to give results on the overall group of 58 072 patients in a weighted analysis.
For the primary analysis, in-hospital mortality was significantly higher among those who had off-pump CABG, at 2.4% vs 1.8% for those who underwent on-pump surgery (odds ratio 1.34; p=0.02). Mid-term survival, with a median follow-up of 2.9 years, was also significantly worse among those who had off-pump surgery (HR 1.28 for mid-term mortality; p=0.002) compared with on-pump.
"In simple terms, it accounts for 108 extra deaths for the off-pump group," Pagano noted.
In the weighted analysis, results were similar, with a 35% increased risk of in-hospital death among those in the off-pump CABG group (p=0.001) and 27% greater mid-term mortality (p<0.0001) compared with on-pump.
But Volume of off-Pump by Institution Modifies the Results
The researchers did find, however, that analyzing the mortality according to volume of off-pump procedures performed per institution modified the results somewhat, at least for short-term, in-hospital deaths.
The median rate of off-pump procedures for institutions in the study was 3.8%, and they divided hospitals into high-performing off-pump CABG centers (>3.8%) and low-performing ones.
For in-hospital deaths, there was no difference in mortality between off-pump and on-pump CABG when the analysis was confined to high-volume centers.
But for mid-term survival, there was still an advantage of on-pump even when it was compared only with off-pump performed at higher-volume institutions, Pagano noted. The hazard ratio was 1.25 for mortality among off-pump patients operated on at such institutions compared with on-pump (p=0.007), "and there was no interaction throughout the data set between the volume of off-pump and the mid-term survival," he said.
But Taggart reiterated his concerns about how the analysis was done. "They propensity-matched their patients very well, but the one thing it doesn't tell you is about the relative experience of the people who were doing these operations," he commented to heartwire . "If it includes surgeons who had a go at doing off-pump surgery and got bad results, they are still included, so this is an institutional analysis as opposed to a per-surgeon analysis, and there is a real danger to that."
But Pagano said it wasn't possible with this database to perform this analysis, and in any case it would suffer from confounding by indication. And while he acknowledged that there is a learning curve for off-pump CABG, he said that in the UK, generally, "in the high-volume units, you mostly have surgeons who do most of their work off-pump."
Taggart told heartwire if the researchers could not perform the analysis in the way he suggested "they shouldn't be making those conclusions."
No Benefits of off-Pump CABG Seen in Any Other Subgroups
Pagano and colleagues also performed a number of subgroup analyses. For those with renal dysfunction, there were too few events to determine whether there was any advantage of one operative procedure over the other. But for those with a previous neurological disorder, who tended to have worse outcomes whichever form of surgery they underwent, "the disadvantage of off-pump did not disappear," he pointed out.
The researchers also went on to pool the results of all other known randomized trial data on this subject in which patients were followed up for at least one year, including the ROOBY trial. These pooled data were remarkably similar to the new UK findings, said Pagano, with a 34% increased risk of in-hospital mortality and a 28% higher risk of mid-term death with off-pump surgery compared with on-pump CABG. "This is where we sit. We are in line with these randomized trials," he observed.
Let's Agree to Disagree
Further discussion after the presentation centered on the fact that off-pump CABG "is a whole family of approaches," according to one attendee, who added to Pagano, "You have not defined your primary variable correctly, and therefore the whole study carries a major flaw."
Pagano said they would have to agree to disagree. "There are no data about the type of off-pump surgeries on survival. While I accept there are many ways of doing OP-CAB, we still don't know the long-term safety, and the purpose of this trial was to generate some reflection."
Meanwhile Dr Teresa Kieser (University of Calgary, AB) said she wondered whether some of the findings could have be driven by graft failure and therefore whether an assessment by the use or lack of use of certain graft assessment techniques would have provided further information.
"We cannot answer this question, since we don't have the data available," said Pagano. "There may be centers and techniques and ways to make sure [off-pump CABG] is safe, but what we are saying here is that actually that's not been tested yet and there are no answers. I wouldn't like to speculate on data I haven't got."
Off-Pump CABG the Norm in Some Asian Centers
As an interesting aside, Pepper told heartwire that in some countries, particularly in Asia, there is an economic advantage to using off-pump CABG.
"There, it is significantly cheaper to use off-pump, and so you will find in, for example, Delhi, Chennai, or Mumbai--the big Indian centers--very large numbers of patients being done off-pump because it's economically sensible. They are very talented surgeons, and they can get comparable results. It's a different economic climate from Europe," he explained.
Asked whether he thought Asian centers would reconsider their high use of off-pump CABG if it is ultimately shown to have no advantage over on-pump CABG or even to have worse outcomes, he commented, "I don't know! Probably not; we'll have to wait and see."
Pagano reports no conflicts of interest. Pagano reports no conflicts of interest.
Off-pump CABG "was associated with a 34% increased risk of in-hospital mortality overall and a 28% reduced risk of mid-term survival, out to almost three years, compared with on-pump," Pagano noted. These new data are in line with findings from smaller, randomized studies, says Pagano, adding, "It's a good snapshot of contemporaneous practice in the UK."
But others criticized the study. Dr David Taggart (University of Oxford, UK), said, "The fundamental flaw in what is potentially a dangerous interpretation is that this was done at an institutional level, and that covers many sins, including surgeons who had attempted off-pump surgery--it didn't work out for them, and they abandoned it. I would want to see these data reanalyzed looking at what happened in the outcomes of surgeons who are high-volume users of off-pump surgery. We need propensity-matching of the operators. That's my concern in the interpretation of these data."
Other attendees pointed out further drawbacks, including the fact that off-pump CABG (also known as OP-CAB) is a whole family of approaches, and the way it was performed in this study was not properly defined. In addition, some of the findings could have been driven by graft failure, and there was no allowance for this in the study design.
Asked to comment on the results, Dr John Pepper (Imperial College London, UK) explained to heartwire there have been "many subtle improvements in heart-lung machines" since off-pump CABG first emerged, and now "the pump is pretty safe. Plus you have to remember that many CABG operations are done in conjunction with valve procedures, which are done on a heart-lung bypass machine."
With this background, "My opinion is that off-pump and on-pump are similar," says Pepper. "Off-pump is not significantly better than on-pump, but in experienced hands, in experienced units, they can achieve extremely good results with off-pump. I think the message is that off-pump is only for a group of people who are very keen on doing it, who have trained themselves obsessively, and then they can get results that are comparable to, but probably not much better than, on-pump."
Pepper added that the results of an ongoing large Canadian randomized clinical trial of off-pump vs on-pump surgery in around 4000 patients are eagerly awaited to try to discern whether there truly is a difference between these two approaches in terms of outcomes.
Compared With on-Pump Surgery, 108 Extra Deaths on off-Pump Group
When off-pump CABG first emerged as a surgical technique, it was thought it would reduce or eliminate the neuropsychological risks of putting a patient on a heart-lung bypass machine, avoiding or reducing so-called "pump head." Pagano explained that off-pump CABG may improve outcomes of selected groups of patients, and "there are a number of small randomized trials that show superiority in some subgroups." But "landmark randomized trials to address short-term and long-term mortality are still not available," he noted.
In the new analysis, he and his colleagues examined data on patients undergoing isolated, first-time CABG between March 2004 and April 2009, obtained from the national database for cardiothoracic surgery of Great Britain and Ireland. Mortality data were extracted from the UK Office for National Statistics. While observational data are no substitute for randomized controlled trials, "they still have a role to provide the best available evidence for practice to date," Pagano observed.
The overall patient population consisted of 58 072 patients, of whom 14% (n=9590) underwent off-pump CABG and 86% (n=48 482) had on-pump surgery. Of these, 6374 off-pump patients were propensity-matched with 6374 on-pump patients for the primary analysis. To further reinforce the main analysis, the propensity score was also used as a covariant to give results on the overall group of 58 072 patients in a weighted analysis.
For the primary analysis, in-hospital mortality was significantly higher among those who had off-pump CABG, at 2.4% vs 1.8% for those who underwent on-pump surgery (odds ratio 1.34; p=0.02). Mid-term survival, with a median follow-up of 2.9 years, was also significantly worse among those who had off-pump surgery (HR 1.28 for mid-term mortality; p=0.002) compared with on-pump.
"In simple terms, it accounts for 108 extra deaths for the off-pump group," Pagano noted.
In the weighted analysis, results were similar, with a 35% increased risk of in-hospital death among those in the off-pump CABG group (p=0.001) and 27% greater mid-term mortality (p<0.0001) compared with on-pump.
But Volume of off-Pump by Institution Modifies the Results
The researchers did find, however, that analyzing the mortality according to volume of off-pump procedures performed per institution modified the results somewhat, at least for short-term, in-hospital deaths.
The median rate of off-pump procedures for institutions in the study was 3.8%, and they divided hospitals into high-performing off-pump CABG centers (>3.8%) and low-performing ones.
For in-hospital deaths, there was no difference in mortality between off-pump and on-pump CABG when the analysis was confined to high-volume centers.
But for mid-term survival, there was still an advantage of on-pump even when it was compared only with off-pump performed at higher-volume institutions, Pagano noted. The hazard ratio was 1.25 for mortality among off-pump patients operated on at such institutions compared with on-pump (p=0.007), "and there was no interaction throughout the data set between the volume of off-pump and the mid-term survival," he said.
But Taggart reiterated his concerns about how the analysis was done. "They propensity-matched their patients very well, but the one thing it doesn't tell you is about the relative experience of the people who were doing these operations," he commented to heartwire . "If it includes surgeons who had a go at doing off-pump surgery and got bad results, they are still included, so this is an institutional analysis as opposed to a per-surgeon analysis, and there is a real danger to that."
But Pagano said it wasn't possible with this database to perform this analysis, and in any case it would suffer from confounding by indication. And while he acknowledged that there is a learning curve for off-pump CABG, he said that in the UK, generally, "in the high-volume units, you mostly have surgeons who do most of their work off-pump."
Taggart told heartwire if the researchers could not perform the analysis in the way he suggested "they shouldn't be making those conclusions."
No Benefits of off-Pump CABG Seen in Any Other Subgroups
Pagano and colleagues also performed a number of subgroup analyses. For those with renal dysfunction, there were too few events to determine whether there was any advantage of one operative procedure over the other. But for those with a previous neurological disorder, who tended to have worse outcomes whichever form of surgery they underwent, "the disadvantage of off-pump did not disappear," he pointed out.
The researchers also went on to pool the results of all other known randomized trial data on this subject in which patients were followed up for at least one year, including the ROOBY trial. These pooled data were remarkably similar to the new UK findings, said Pagano, with a 34% increased risk of in-hospital mortality and a 28% higher risk of mid-term death with off-pump surgery compared with on-pump CABG. "This is where we sit. We are in line with these randomized trials," he observed.
Let's Agree to Disagree
Further discussion after the presentation centered on the fact that off-pump CABG "is a whole family of approaches," according to one attendee, who added to Pagano, "You have not defined your primary variable correctly, and therefore the whole study carries a major flaw."
Pagano said they would have to agree to disagree. "There are no data about the type of off-pump surgeries on survival. While I accept there are many ways of doing OP-CAB, we still don't know the long-term safety, and the purpose of this trial was to generate some reflection."
Meanwhile Dr Teresa Kieser (University of Calgary, AB) said she wondered whether some of the findings could have be driven by graft failure and therefore whether an assessment by the use or lack of use of certain graft assessment techniques would have provided further information.
"We cannot answer this question, since we don't have the data available," said Pagano. "There may be centers and techniques and ways to make sure [off-pump CABG] is safe, but what we are saying here is that actually that's not been tested yet and there are no answers. I wouldn't like to speculate on data I haven't got."
Off-Pump CABG the Norm in Some Asian Centers
As an interesting aside, Pepper told heartwire that in some countries, particularly in Asia, there is an economic advantage to using off-pump CABG.
"There, it is significantly cheaper to use off-pump, and so you will find in, for example, Delhi, Chennai, or Mumbai--the big Indian centers--very large numbers of patients being done off-pump because it's economically sensible. They are very talented surgeons, and they can get comparable results. It's a different economic climate from Europe," he explained.
Asked whether he thought Asian centers would reconsider their high use of off-pump CABG if it is ultimately shown to have no advantage over on-pump CABG or even to have worse outcomes, he commented, "I don't know! Probably not; we'll have to wait and see."
Pagano reports no conflicts of interest. Pagano reports no conflicts of interest.
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου