Ann Thorac Surg 2000;70:S38-S42
© 2000 The Society of Thoracic Surgeons
Coronary artery bypass graft in abciximab-treated patients
Leroy A. LeNarz, MDa,1
a Lilly Corporate Center, Indianapolis, Indiana, USA
Address reprint requests to Dr LeNarz, Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN 46285
e-mail: lenarz_leroy_a{at}lilly.com
Presented at the "Managing the Patient Receiving Platelet Inhibitors in Cardiac Surgery" Roundtable Discussion, San Antonio, TX, Jan 22–23, 1999.
 |
Abstract
|
|---|
In an assessment of available literature regarding appropriate treatment algorithms for patients who fail cardiac interventions and require coronary bypass grafting with abciximab "on board," few surgeons reported significant variation in bleeding or other issues. There are many problems with postmarketing data collection, and such data would lack rigor and randomization. Prospective collection of data would be extremely difficult. Few patients go to surgery nowadays, and it is generally believed that because of the wealth of data supporting glycoprotein (GP) IIb/IIIa inhibition associated with percutaneous intervention, it would be unethical to subject patients to a placebo control in the catheterization laboratory. For those reasons, we returned to the clinical trials database and did a retrospective analysis of EPILOG and EPISTENT, the two studies that most appropriately reflect state-of-the-art intervention and abciximab use. We omitted EPIC, the original registration trial of high-risk angioplasty patients, because this study no longer reflects current recommendations regarding heparinization, abciximab administration, or groin care.
|
Introduction
|
|---|
It is proper to begin this discussion with background information. By training and prior practice, I am a board-certified cardiovascular surgeon. But because of a totally denervating injury to my right ulnar nerve, I abandoned surgery in 1992. Currently, I am a regulatory advisor for cardiovascular products for Eli Lilly and Company.
In that role, I have had the opportunity to field questions from sales personnel, surgeons, cardiologists, and perfusionists regarding appropriate treatment algorithms for patients who fail cardiac interventions and require coronary bypass grafting with abciximab "on board."
A review of available information in the clinical trials database revealed data collected via a method that is customary for cardiology trials but not for surgical hemostatic trials. Cardiology trials make reference to thrombolysis in myocardial infarction (TIMI) transfusion criteria, a measure of blood loss classified as major or minor TIMI bleeding. TIMI major blood loss is defined as intracranial hemorrhage or blood loss resulting in a drop in hemoglobin by greater than 5 g/dL. Criteria for minor bleeding include gross hematuria, hematemesis, a decrease in hemoglobin of 3 to 5 g/dL associated with observed bleeding, or a decrease by 4 to 5 g/dL if no active site of bleeding is identified.
In an assessment of available literature, few surgeons reported significant variation in bleeding or other issues. There are many problems with postmarketing data collection, and such data would lack rigor and randomization. Prospective collection of data would be extremely difficult. Few patients go to surgery nowadays, and it is generally believed that because of the wealth of data supporting glycoprotein (GP) IIb/IIIa inhibition associated with percutaneous intervention, it would be unethical to subject patients to a placebo control in the catheterization laboratory.
|
EPILOG and EPISTENT: state-of-the-art intervention
|
|---|
For all the above reasons, we elected to return to the clinical trials database and do a retrospective analysis of the EPILOG and EPISTENT clinical trials. We omitted EPIC, the original registration trial of high-risk angioplasty patients, because this study no longer reflects state-of-the-art heparinization, abciximab administration, or groin care. Application of the lessons learned in EPIC regarding weight adjustment of heparin and abciximab lowered groin-bleeding rates to placebo event rates in EPILOG. Stent usage has changed intervention and urgency for bypass. Therefore, our consultants advised us, and we agreed, that review of these two studies most appropriately reflects state-of-the-art intervention and abciximab usage.
George Despotis, MD, helped design a case report form used by the Cleveland Clinic to collect data not previously reported on the cardiology trial case report form. Additional information to be collected included chest tube blood loss, heparin dosing, activated clotting time (ACT) measurements, and information to allow risk stratification between groups.
|
EPILOG/EPISTENT data analyses
|
|---|
Table 1 gives the number of patients who went to surgery after failed intervention during EPILOG and EPISTENT. After abciximab treatment, approximately 1% fewer failed angioplasty patients will need to be sent to surgery. This fact should please surgeons, because these patients are at higher risk for perioperative events.
I should point out that there have been other researchers who have discussed urgent revascularization rates with regard to these trials. "Urgent" by trial definition refers to the emergent need for intervention and does not reflect the recent administration of abciximab. Once we recognized this, we made it a point to note the timing of abciximab discontinuance and coming off pump in the data collection, because this would be the point at which a surgeon would first have to deal with potential bleeding. We also noted the times "from off pump to closure," another potential measure of microvascular bleeding.
As previously mentioned, we collected information regarding heparin usage and ACT. We know from the sales force that there are cardiologists who recommend that surgeons decrease heparin on the pump in the operating room, because we have been able to do this in the catheter lab with less bleeding in EPILOG. We could spend all day discussing the safety issues of such recommendations and the polarity that exists in the surgical/perfusion community regarding heparin dosing, ACT measurements, and targets for bypass. We want to avoid that discussion. However, we do know that abciximab may elevate the ACT by 35 seconds in the range used for catheter lab procedures. With hemodilution, cooling, and ACT levels targeted for bypass, one can only speculate as to abciximab effects. We will present heparin dosing data but cannot make a recommendation to allow alteration of heparin dosing to targeted ACTs on the pump.
The data to be presented are the early results from our retrospective look. Please be aware, however, that data continue to come in and that final results will be published in manuscript form later this year. The rationale for presenting these data with other cuts at the clinical trial database is to confirm that our message to the surgical, perfusion, and cardiac anesthesia groups charged with the care of these patients is appropriate. We also wish to determine if any additional data need to be generated either from the clinical trials database or from proposed additional studies.
Because so few patients go to surgery from any one site during clinical trials, and because most sites do not regard this data collection as a priority, it is not always easy to convince study site coordinators to collect charts and fill out an additional two-page case report form. The Cleveland Clinic Coordinating Center has done a very good job of prompting site study coordinators to make reporting complete so that we would have as many patients as possible available for review in the final manuscript. We now anticipate that there will be only a few missing charts, an excellent result considering the EPILOG sites were closed 3 years ago.
Let me restate that we are dealing with two databases: (1) the original data associated with the trial, and (2) data obtained from the retrospective coronary artery bypass graft (CABG) two-page case report form. The original database contained demographic data that allowed a look at the profiles of patients going on to surgery during initial hospitalization. We can state with certainty that the demographics (Table 2) were not weighted or slanted in favor of abciximab. The abciximab-treated patients include a few more females than males and more with diabetes than without. There is no significant difference as to age, weight, race, or body surface area.
Approximately 1 in 4 from placebo and treated groups went to the operating room with an intraaortic balloon pump (IABP) in position, reflecting their degree of hemodynamic compromise and the true emergent nature of their situation (Table 3).
We will now look at the data relative to ACT and heparin (Tables 4, 5). One of the frustrations has been the difficulty in discerning from the charts what type of machine was used to assess ACT. Therefore, these data leave in question the devices used to target some of the ACT ranges discussed.
It should be noted that the total heparin dose used was lower in the abciximab-treated patients who achieved higher ACTs during cardiopulmonary bypass.
One could surmise that the ACT-targeted range is achieved with less heparin plus abciximab, but again, lack of additional data regarding measurement devices plotted against the duration of heparin with cooling and hemodilution creates problems (Table 6). It may be appropriate to assume that ongoing monitoring of the ACT on bypass led to fewer heparin boluses to maintain desired ACT ranges by the perfusionist and anesthesiologist. Because I represent the sponsor, it would, at this time, be inappropriate for me to make recommendations as to a specific targeted ACT for bypass, without additional information regarding abciximab, heparin, ACT interaction, and the effects of cooling and hemodilution.
Total protamine dose does not differ, which may indicate that significant effort was not made to stop microvascular bleeding with additional protamine in the abciximab-treated group (I believe that, right or wrong, there is a tendency to give more protamine in the postpump period if patients continue to ooze).
After discontinuation of abciximab infusion, platelet function returns gradually to normal. Bleeding time returned to 12 minutes within 12 hours postinfusion in 75% of a previously studied group of patients. Platelet aggregation is measured using various stimulating agents. In response to 20 µM adenosine diphosphate (ADP), ex vivo platelet aggregation returns to 50% of baseline within 24 hours in 62% of patients. Platelet transfusion allows for a quicker return of platelet function tests to normal ranges because of abciximab reversibility.
Because of this known reversibility, we bracketed the patients at 0 to 6 hours, 6 to 12 hours, and >24 hours, to delineate the effect of redistribution of abciximab across platelets and to link the timing of surgery to discontinuance of the drug. It should be noted that the assumption of higher bleeding rates within the first 12 hours after cessation of abciximab could result in delaying surgery in a patient population that may be in the greatest need of urgent revascularization for salvage.
It appears that a high percentage of patients (both placebo- and abciximab-treated) go to the operating room within 6 hours after intervention (Table 7), which is not surprising. There is a difference of less than 12 minutes between the groups from off pump to closure. This implies a not too significant amount of time spent "mopping up." Regarding chest tube drainage, there is no significant difference at 6 hours or for total loss measured, whether reported as median or mean chest tube loss (Table 9).
Reexploration rate was higher in the abciximab arm (Table 10), although exact findings at the time of reexploration are hard to reconstruct in this type of late retrospective analysis. The number of patients reexplored for diffuse ooze is similar. However, other findings were not noted or are not available (eg, rate of vein graft or inferior mesenteric artery [IMA] branch bleeding; occurrence of right or left ventricular failure).
One of the other issues Dr Despotis wanted to explore was the quantity of cryoprecipitate required. The feeling was that when there is bleeding that is not usually controllable, surgeons and anesthesiologists might resort to "shotgun" use of cryoprecipitate as an intravenous treatment, or as a part of tissue glue. As for the number of cryoprecipitate-transfused patients, it is basically the same in each group (Table 11). There was less total volume of cryoprecipitate administered in the abciximab group.
Overall, more abciximab-treated patients than control patients received platelets (Table 12), and more abciximab-treated patients (67%) than control patients (43%) received platelets when operated on within 12 hours of drug administration (Table 13). The treatment for patients going for CABG was unblinded, possibly prompting unnecessary platelet administration (not discernible in this type of retrospective analysis). Another way to look at this information is that 33% of abciximab-treated patients operated on within the first 12 hours did not receive platelet transfusion. It is apparent that both platelet and red blood cell (RBC) usage is higher in the control group than in the abciximab-treated patients going to surgery within 12 hours of intervention. I suspect that the latter patients, compared with other groups, were probably sicker and more hemodynamically compromised with an IABP in position, which is also known to cause platelet dysfunction. Additional vascular access site issues associated with intraaortic balloon also increase the opportunity for transfusion.
If one breaks down the studies and looks at EPISTENT, the highest percentages of transfusion of both RBCs and platelets were in the abciximab plus stent arm (Table 14). These patients received aspirin, ticlopidine, and abciximab and had the greatest treatment (lowest event) effect. It may be that the nonreversible effects of other drugs contribute to greater bleeding when combined with abciximab, or there may be a synergism of all three platelet inhibitors that results in greater bleeding, or it may be that these patients are the "real bleeders."
This brings up an important point regarding abciximab. Platelet transfusion allows for redistribution of abciximab, less receptor occupancy, and quicker return to more normal platelet aggregation. It must be emphasized again that the effects of abciximab are reversible.
The events analysis demonstrated that 4 of 38 control patients versus 2 of 44 abciximab-treated patients died within 30 days of surgery (Table 15). Although not statistically significant, a 5% mortality in a group of patients with recent failed angioplasty (21% of whom had an IABP in position) is not inconsistent with other reports of mortality rates in emergent situations. I am aware of one single-institution series in Boston in the mid-1990s with a 30-day mortality rate of 13% (Table 16).
|
Conclusion
|
|---|
Our series reflects the positive impact of abciximab plus stent: sending fewer patients to surgery (Table 17) with less than historical or placebo-controlled mortality.
Although the numbers are too small to allow comparison across groups for combined event rates of myocardial infarction and death, differences are not statistically significant, and certainly the trend does not reflect adversely on abciximab. This information, combined with the reduced number of seriously ill patients requiring surgery after failed angioplasty, should help convince surgeons that it is safe to operate whenever surgery is clinically mandated. (Table 8)
The protocol for management of abciximab-treated patients undergoing urgent CABG (Table 18) is to discontinue the abciximab infusion; the ACT should be checked in the operating room, without adjusting heparin or reversing heparin with protamine. It is important to remember that the effect of abciximab will wear off. If one has a clinical bleeder, platelet transfusion should be considered as an antidote after bypass and postheparin neutralization. Remember, platelet transfusion will reverse the hemostatic effect of abciximab.
View this table:
[in this window]
[in a new window]
|
Table 18. Management of Patients Treated With Abciximab Undergoing Urgent Coronary Artery Bypass Graft: Recommendations
|
|
|
Footnotes
|
|---|
1 Doctor LeNarz is an employee of Eli Lilly and Company, Indianapolis, IN. 
This article has been cited by other articles:
|
|
|
|
 
J. Barnard and R. Millner
A review of topical hemostatic agents for use in cardiac surgery.
Ann. Thorac. Surg.,
October 1, 2009;
88(4):
1377 - 1383.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. H. Hongo, J. Ley, S. E. Dick, and R. R. Yee
The effect of clopidogrel incombination with aspirin whengiven before coronary artery bypass grafting
J. Am. Coll. Cardiol.,
July 17, 2002;
40(2):
231 - 237.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
