| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Click here to read this article as a CME activity
Ann Thorac Surg 2007;83:912-920
© 2007 The Society of Thoracic Surgeons
a Children’s Hospital Boston, Boston, Massachusetts
b St. Louis Children’s Hospital, St. Louis, Missouri
c Children’s Hospital Pittsburgh, Pittsburgh, Pennsylvania
Accepted for publication September 18, 2006.
* Address correspondence to Dr Pigula, Department of Cardiac Surgery, Bader 274, Boston Children’s Hospital, Boston MA 02115 (Email: cbaird{at}sanger-clinic.com).
Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30–Feb 1, 2006.
| Pediatric cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Methods: A retrospective review was conducted of 604 consecutive pediatric ECMO patients at a single institution between 1980 and 2001. Multiple logistic regressions were used to assess the impact on survival of ACT, heparin, age, weight, diagnosis, and previous surgery.
Results: There were 349 survivors (57.8%), and 255 (42.2%) nonsurvivors. Mean hours on ECMO were 182 ± 134 (range, 3 to 957 hours), mean ACT was 227 ± 50 seconds (range, 158 to 620 seconds), and the mean hourly heparin dose was 45 ± 21 U/kg (range, 6 to 134 U/kg). Regression analysis indicated that increased heparin administration was predictive of survival (p < 0.0001), independent of all other variables. The ACT was not a predictor of survival (p = 0.096). Although previous surgery was independently associated with an increased likelihood of ECMO death (p < 0.001), increased heparin administration again exerted a survival advantage (p = 0.012).
Conclusions: Adherence to the recommended ACT of 180 to 220 seconds in pediatric ECMO patients may result in inadequate anticoagulation. Survival is improved by increased heparin administration independent of the ACT. The ACT may be too insensitive to maintain adequate long-term systematic anticoagulation, and other methods, such as heparin levels or functional parameters such as anti-Factor Xa activity or thrombin generation, should be investigated.
Pediatric extracorporeal membrane oxygenation (ECMO) has been used to treat thousands of neonates and children with severe respiratory distress and congenital heart disease. Despite many technical advances and the refinement in management techniques, we continue to be faced with the challenge of balancing anticoagulation. An ideal anticoagulant would avoid thrombotic and hemorrhagic complications, be readily titratable, and easy to control.
Unfractionated heparin (UFH) has been available commercially for more than half a century and is the most widely used anticoagulant. It quickly binds to and activates antithrombin, which then inhibits activated factors in the intrinsic clotting cascade. Its short half-life and the fact that its action can be reversed readily with protamine make it an almost ideal antithrombotic agent. The ability to accurately monitor heparin anticoagulation has been difficult, however.
In 1966, owing to the lack of specific and sensitive tests for determining coagulation, activated coagulation time (ACT) of whole blood was created. Two milliliters of venous blood were placed into a warmed tube containing diatomaceous earth, which was tilted and observed for clot formation [1]. It has subsequently been applied to monitor heparin therapy in a multitude of settings.
Soon after its introduction, Bull and colleagues [1] began investigating ACT protocols for heparin therapy used during extracorporeal circulation for open heart surgery and made several important fundamental observations. They first observed that similar ACTs in different patients resulted in inadequate heparinization or neutralization of heparin by protamine. Next, they recognized that the amount of heparin required to produce an arbitrary prolongation of ACT varied threefold between patients. Finally, the rapidity with which heparin disappears from the blood may vary fourfold. To correct for the inherent inadequacies of the ACT test, dose-response curves relating heparin dosage and effective ACT were achieved, making it possible to maintain safer ranges of anticoagulation during extracorporeal circulatory support [2, 3].
In a recent study by Graves and colleagues [4] surveying all active neonatal ECMO programs in the United States, ACT was the predominant anticoagulation test, but there was no consensus on the type of heparin, frequency of ACT testing, or methods for dealing with abnormal ACT results. Some programs had a minimum heparin dose despite ACT levels, some strictly followed ACT results, and others discontinued heparin for various reasons [4]. Furthermore, previous studies suggest that an ACT rate of between 180 to 220 seconds results in adequate anticoagulation during extracorporeal circulation [5]. The purpose of this study was to determine the relationships between heparin dose, ACT, and survival.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
General Extracorporeal Membrane Oxygenation Protocols
Retrospectively reviewed were 604 consecutive pediatric ECMO patients at Children’s Hospital of Pittsburgh from 1980 to 2001. ECMO was initiated for respiratory failure, cardiac failure, and sepsis when all other options failed. The causes for respiratory failure in these patients were meconium aspiration, congenital diaphragmatic hernia, primary pulmonary hypertension, and pneumonia. The cardiac patients had undergone primary cardiac repair, orthotopic heart transplant, and bridge to transplant.
Venoarterial ECMO was used in all patients with size-appropriate cannulas. For all primary respiratory failure patients, the right internal jugular vein was cannulated, and postcardiotomy patients were routinely cannulated through the right atrium and aorta. Early in the experience, all children were supported with roller pump ECMO, whereas during the last half of the experience, children weighing more than 15 kg were supported with centrifugal pumps. ECMO flows were initiated at 100 to 150 ml/(kg · min).
Anticoagulation Methods
Patients received an unfractionated heparin (UFH) loading dose of 100 U/kg immediately before cannulation, with an additional 100 U/kg in the blood prime. Initial ACTs were checked immediately after cannulation and allowed to drift down to the target ACT of 180 to 220 seconds. The heparin dose was then adjusted based on ACT levels as well as the perceived bleeding risk. Whole blood ACTs were determined hourly for the first 48 hours, every 4 hours after 48 hours, and as necessary thereafter. From 1980 through 1990, the ACT was determined with Celite diatomaceous earth (Celite Corporation, Goleta, CA) as the primary reagent, and the Hemochron JR (International Technidyne, Edison, NJ), which uses kaolin/silica reagents, was used between 1990 and 2001. Although the "target" ACT was 180 to 220 seconds, modifications were made at the discretion of attending physicians, and ACTs were sometimes modified to reflect the clinical situation such as bleeding or hypercoagulability and thrombus formation in the circuit, which often required replacement of the circuit.
Statistical Analysis
ACT and heparin doses reflect an average over the course of 24 hours and 7 days. Continuous data, including ECMO duration, heparin dose, and ACTs, were tested for normality using the Kolmogorov-Smirnov statistic, and no significant departures were detected. Therefore, these variables were compared between survivors and nonsurvivors using the two-sample Student t test and stratified according to previous surgery status.
The Pearson product-moment correlation coefficient (r) was used to summarize the association between heparin dose and ACTs.
Multiple stepwise logistic regressions were applied to determine the independent predictors of patient survival, considering ACT level, heparin dose, ECMO duration, and previous surgery status as variables in the multivariate analysis. Based on the plotted empirical data, model fitting for determining the most accurate functional relationship between heparin dose and survival included testing whether a quadratic fit captured the relationship better than a linear fit [6].
The likelihood ratio test was used as the measure of significance, and odds ratios with 95% confidence intervals (CI) were calculated for significant predictors in the final model (Table 1) [7]. Maximum likelihood estimation was used to derive the predicted probability of patient survival based on the regression coefficients for heparin dose and ACT (Table 2) [8]. Statistical analysis was performed with the software package SPSS 14.0 (SPSS Inc, Chicago, IL). Two-tailed values of p < 0.05 were considered statistically significant [6].
|
|
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Death occurred in 65% of cardiac patients, 36% of respiratory patients, 50% of sepsis patients, and 50% of other patients.
For all patients (survivors and nonsurvivors) in this study, median patient age was 3 days (interquartile range; 1 to 52 days). The mean duration on ECMO was 182 ± 134 hours (range, 3 to 957 hours). Cardiac patients were on ECMO fewer hours than respiratory patients (129 ± 115 hours versus 195 ± 135 hours; p < 0.05). Among patients with a cardiac diagnosis, there were no significant differences in ECMO duration. Of all cardiac patients, however, survivors had fewer hours on ECMO than nonsurvivors (122 ± 96 hours versus 133 ± 124 hours), although this was not statistically significant (p = 0.65). Similarly, among all respiratory patients, the average time nonsurvivors were on ECMO was longer than survivors (228 ± 173 hours versus 175 ± 102 hours; p < 0.001).
For all patients, mean ACT was 227 ± 50 seconds (range, 158 to 620 seconds) and the mean hourly heparin dose was 45 ± 21 U/kg on ECMO (range, 6 to 134 hours). Significantly less heparin was given to cardiac patients compared with pulmonary patients on ECMO (40 U/[kg · h] versus 46 U/[kg · h]; p < 0.05). Relatively larger percentages of cardiac patients died who had less heparin compared with pulmonary patients.
To determine variations among heparin and ACT, a linear correlation was developed. Only a moderate correlation was found between heparin doses and ACT for all ECMO patients (Pearson r = 0.48; Fig 1). Pearson correlation coefficients between heparin dose and ACT for survivors and nonsurvivors were r = 0.48 and r = 0.42, respectively, each only with a moderate correlation. However, by crude inspection of Figure 1, most of the nonsurvivors were in the left lower quadrant, and most of the survivors were in the right upper quadrant.
|
Results were similar when considering patients with no previous surgery, whereby survivors received an average of 10 U/(kg · hr) more than nonsurvivors (49 ± 19 U/[kg · h] versus 39 ± 23 U/[kg · h]; p < 0.001). Moreover, with respect to patients with previous surgery, although no significant difference was detected in ECMO duration, survivors received significantly more heparin than nonsurvivors (50 ± 25 U/[kg · h] versus 36 ± 19 U/[kg · h]; p = 0.012). Mean ACTs were not significantly different between survivors and nonsurvivors when considering all patients (p = 0.36), patients with no previous surgery (p = 0.65), or those with previous surgery (p = 0.45).
Figure 2 shows this relationship between heparin dose and the predicted probability of survival where the theoretic curve depicts greater protection with an increased amount of heparin usage (p < 0.001) but no statistically significant patient survival benefit based on differences in ACTs (p = 0.36), where the probability of survival is estimated to be 58% across the range of possible values of ACT. These curves were determined based on univariate logistic regression.
|
Independent of ACT, ECMO duration, and previous surgery status, each increase of 10 U/(kg · h) of heparin is predictive of an increase in the odds of survival of 1.27. This can be converted to probability of survival using the simple formula: Probability = Odds/(1 + Odds) = 1.27/2.27 = 56%. Thus, patients have an increased 56% probability of survival for each increase of 10 U/(kg · h) of heparin. In addition, patients with no previous surgery had significantly higher odds of survival, estimated at 3.7 times higher, independent of ECMO duration, ACT level, or amount of heparin used. No significant interaction was found between heparin given and previous surgery (p = 0.48). ACT levels were not found to be significantly predictive of patient survival (p = 0.096), although the p value reflects a trend suggesting some benefit of higher ACT levels—surprisingly, because it would seem from the multivariate results that a longer ACT is detrimental.
There were no significant differences within the first 24 hours between postoperative heparin start times or between patients with and without previous surgery. However, significant differences were found in heparin doses on ECMO between survivors and nonsurvivors (44 ± 20 U/[kg · h] versus 35 ± 22 U/[kg · h]). When evaluating survival outcome at 24 hours and at 7 days, both analyses revealed significantly higher heparin doses among survivors (p < 0.001).
Figure 2 represents empirical histograms describing the percentage of patients in each of 10 intervals of heparin dose for survivors and nonsurvivors. For example, 14% of survivors and 40% of nonsurvivors had heparin levels of less than 30 U/(kg · h) on ECMO. On the other hand, 26% of survivors and 14% of nonsurvivors had heparin doses exceeding 60 U/(kg · h) on ECMO. Superimposed is the theoretic probability curve showing survival according to heparin dose interval. Logistic regression modeling revealed a highly significant increased probability of survival for patients receiving greater heparin doses (p < 0.0001).
For an acceptable heparin dosing of 40 U/(kg · h) or more, the chance of survival was 71%, whereas for less than 40 U/(kg · h), the chance of survival was only 47%. For an acceptable heparin dosing of 50 U/(kg · h) or more, the chance of survival was 72%, whereas for less than 50 U/(kg · h), the chance of survival was only 53%. For an acceptable heparin dosing of 60 U/(kg · h), the chance of survival was 73%, whereas for less than 60 U/(kg · h), the chance of survival was only 57%. This considers the accumulation of all patients from very low heparin doses up to 60 U/(kg · h), at which there is a doubling (relative risk = 2.1) of survival with heparin doses of 60 U/(kg · h) or more. Specifically at each heparin dose of 30, 40, 50, 60, and 70 U/(kg · h), the respective probabilities of survival are 50%, 58%, 64%, 70%, and 75% (Table 3).
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Potential explanations for survival benefit of increased heparin levels could include a decrease in thromboembolic complications such as stroke, and pulmonary and peripheral emboli. Furthermore, potential causes for decreased survival in patients having had previous surgery include increased bleeding, more complex cardiac and pulmonary disease, and increased operative times. It was concerning that deaths were greater and heparin doses were lower in cardiac patients compared with pulmonary patients. Explanations could include preexisting anticoagulation and the postoperative status of these patients. Furthermore, ACT levels did not necessarily correlate with increased heparin doses, and potential explanations for this discrepancy may be patient-related or technique-related, or both, which will be discussed subsequently.
Heparin has been available commercially for more than half a century and has been the most widely used agent for quickly suppressing thrombosis. The short half-life of 90 minutes and its reversibility by protamine make it a convenient and effective anticoagulant for mechanical circulatory support. Although heparin titration based on the ACT has been a very successful strategy, long-term titration of this is drug difficult. ECMO requires long-term systemic anticoagulation, and heparin titration has been extrapolated from acute cardiopulmonary bypass (CPB) regimens.
Anticoagulation requirements among pediatric patients are not identical to adult patients. Initial studies examining anticoagulation on neonatal ECMO patients in lower ACT ranges were clearly suboptimal and inadequate [9]. This could be attributed to numerous factors that alter patient physiology while on ECMO, including activation of the fibrinolytic and inflammatory systems, various coagulation factors, and the intrinsic pathway, as well as platelet activation and adherence to the circuit. Furthermore, the hemostatic system is affected by gestational age, actual age, organ system damage, or an operation while on ECMO, alone or combined.
Hemorrhage secondary to heparin therapy is a principal cause of morbidity and mortality in pediatric patients treated with ECMO. Hirthler and colleagues [10] reported that increased difficulty with control of the ACT, platelet count, or both had occurred in ECMO deaths associated with intracranial hemorrhage. In their study, despite discontinuation of heparin, this manifested as an inability to raise the ACT or decrease the ACT, or both. These patients also had a significantly greater number of changes in the rate of heparin infusion than did matched controls [10].
Increased levels of anticoagulation are, however, necessary to minimize the devastating and potentially fatal thromboembolic potential associated with ECMO [1]. Some of the more obvious consequences of inadequate anticoagulation include thrombus formation, clot in the circuit, and the need to change the circuit more frequently. Thus, ideally, we need an antithrombotic agent that can provide adequate anticoagulation with a reliable test for accurate monitoring.
Despite showing a strong correlation between higher heparin doses and improved survival, we have shown only a moderate positive correlation between heparin and ACTs between all patients (0.47), survivors only (0.48), or nonsurvivors (0.42). The anticoagulant effect of heparin depends upon plasma properties as well as heparin concentrations and differs between subjects. Biologic activity of heparin in different patients may be as short as 30 minutes or as long as 6 hours. Direct monitoring of circulating heparin concentration through bioassays suggests that ACTs or activated partial thromboplastin time (aPTT) differ in their reflection of the state of anticoagulation [11].
Several factors besides serum heparin concentration have been shown to affect the ACT result. These can be divided into patient-related and technically related factors. Patient-related factors include platelet count, fibrinogen, fibrin degradation products, packed red blood cell volume, hemodilution, and antithrombin III levels [11, 12]. Furthermore, determination in complex and sick patients has its own further irregularities. De Waele and colleagues [13] studied the accuracy of the ACT in predicting the level of anticoagulation and the correlation between ACT and aPTT in critically ill patients. Analysis revealed a significant difference in ACT between different levels of anticoagulation, suggesting a poor correlation between the aPTT and ACT in the intensive care unit setting. [13]. Other patient-specific factors, including hemodilution and age-related differences in fibrinogen levels, impact heparin’s anticoagulation efficacy [14].
Although we did not measure any markers for thrombin formation (prothrombin fragment F1.2) or fibrinolysis (D-dimers), it is possible that a consumptive coagulopathy existed in these patients and also contributed to the inadequacy of ACT measurements. To confirm these assumptions, it would be necessary to study the hematocrit and platelet counts, as well as transfusion requirements of red blood cells and platelets; however, this was not possible in this study, which was conducted in a retrospective fashion over 10 years.
Technical factors that may impact ACT accuracy include agitator and timers [15], lack of single sample precision [16], venous versus arterial blood [17], smaller sample volumes [4], temperature, and timing of ACT levels [18]. Many studies [9, 19, 20] support that different ACT analyzers, such as HemoTec, Hemochron, and Hemochron Jr, frequently produce different results. One must therefore compare ACTs from the same analyzer. All methods of ACT measurement have shown significant increases (p < 0.05) in clotting times at hypothermic CPB [21], whereas hyperthermia induces a significant decrease in clotting times [14]. The ACT test is usually performed immediately after blood collection; however, certain situations often occur that delay rapid measurement. Utilizing eight different devices at different time points, Searles and colleagues [22] suggest that to obtain accurate ACTs, they must be performed on blood samples in less than 15 minutes, as was done in our study.
The many potential causes for inaccurate or varying ACT levels make them less than ideal to determine the most effective dose of heparin anticoagulation. This provides a multitude of reasons to support our findings of only a moderate correlation between heparin and ACTs. In minimizing ACT variability, one should perform repeated measurements, use the same machine, larger sample volumes, optimize temperature, and should not alter commercial reagents because they vary widely in their in vitro sensitivities to heparin. Furthermore, to improve the sensitivity of the ACT one should consider modifications of the standard ACT such as the Actalyke (Helena Laboratories, Beaumont, TX) modification [23].
Alternatives methods to measure heparin induced anticoagulation should consider functional tests such as aPTT, anti Xa, thromboelastography, or heparin/protamine titration methods (Hepcon, Medtronic Perfusion Systems, Minneapolis, MN) or nonfunctional methods such as direct serum heparin levels. It should be remembered that with any laboratory test the sensitivity of critical reagents may vary, affecting the results. Consequently, it is extremely important for each laboratory to carefully identify the lower and upper limits of the normal range.
The continuous coagulation profiles of TEG (Haemoscope Corp, Niles IL) provide insight from the initiation of hemostasis through final clot formation incorporating anticoagulation factors other than thrombin. TEG capabilities allow one to monitor platelet function, clotting factors, and fibrinolysis. Specifically, with regard to neonatal ECMO, Zavadil and colleagues [24] suggest factors other than heparin contribute to the derangement in hemostasis, and the interpretation of TEG data is invaluable. Applying an algorithm including TEG enables physicians to achieve a more accurate reflection of the in vivo physiology of anticoagulation [25]. A novel method utilizing a TEG assay with tissue factor and kaolin (TEG TF/K) more rapidly and accurately monitors heparin anticoagulation [26].
ECMO and CPB lead to fulminant activation of the hemostatic-inflammatory system. Owings and colleagues [27] have shown that children undergoing CPB with heparin dosing adjusted to optimize ACT manifest inadequate levels of anticoagulation secondary to lowered fibrinogen levels. They suggest using heparin management tests that are less dependent on fibrinogen levels than ACT [27]. The Hepcon/HMS system automatically provides the ACT and a whole blood heparin concentration. It also provides the adequate protamine dose by titration of protamine to heparin. Clinical experience with the Hepcon suggests that this method of heparin management during CPB reduces hemostatic activation and the inflammatory response.
In a randomized prospective study, Koster and colleagues [28] compared the influence of anticoagulation with a system based on heparin concentration (Hepcon HMS) with that of ACT-based management on the activation of the hemostatic-inflammatory system during CPB. After CPB, there were significantly higher concentrations for heparin and a significant reduction in thrombin generation, D-dimers, and neutrophil elastase, and a trend toward lower ß-thromboglobulin, C5b-9, and soluble P-selectin in the Hepcon HMS group. There were no differences in the post-CPB values for platelet count, adenosine diphosphate-stimulated platelet aggregation, antithrombin III, soluble fibrin, factor XIIa, or postoperative blood loss. Thus, compared with ACT based management, heparin concentration–based anticoagulation management during CPB leads to a significant reduction of thrombin generation, fibrinolysis, and neutrophil activation, whereas there is no difference in the effect on platelet activation [28].
When the Hepcon/HMS device was used in 45 patients undergoing congenital heart repair with CPB, it was concluded from the variation in heparin dose responses that the conventional way of heparin administration according to body weight alone may cause inadequacy of anticoagulation during CPB [29]. Thus, just as the use of this protamine titration method for heparin dosing in pediatric CPB patients, it may also provide a more effective and accurate blood heparin level in pediatric ECMO patients.
Although heparin remains the traditional method of acute anticoagulation for mechanical circulatory support, alternatives to convention heparin anticoagulation are becoming available. New low-molecular-weight heparins (LMWH) may provide an alternative to UFH. Roschitz and colleagues [30] have shown that the dose to plasma activity of LMWH was more consistent than the UFH group in the pediatric cardiac catheterization lab. They concluded that LMWH was equally efficacious to UFH bolus administration, as determined by plasma levels and markers of clotting activation. Furthermore, in contrast to UFH bolus, no further monitoring was necessary after the application of LMWH during cardiac catheterization owing to a consistent dose to plasma activity [30]. The anti-Xa assay using chromogenic substrates is the most specific and valid one for monitoring LMWH therapy [31].
Direct thrombin inhibition by direct thrombin inhibitors such as argatroban, hirudin and its analogues have become common alternatives to the anticoagulation management of patients unable to tolerate heparin, but their use in long-term mechanical support has not been prospectively examined [32–34].
In summary, the major finding of the study was a strong correlation between heparin administration and survival among all patients as well as cardiac and respiratory groups, separately. Although a modest correlation was found between ACT and heparin dose, increased ACT was not associated with improved survival. Heparin-based anticoagulation has proven very effective for relatively brief periods of CPB, but heparin anticoagulation contributes to the significant morbidity and mortality associated with longer-term mechanical circulatory support. These results suggest that our current methods of heparin titration for pediatric ECMO support are inadequate. Improved monitoring techniques, or alternative agents, or both, should be explored in an effort to improve the outcome of patients supported by mechanical means.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
What I think you have concluded from the study is that (A) we need to give more heparin, and (B) ACT isn’t a good test. And what I would suggest is if we look at it in a little bit different way, what we need isn’t necessarily more heparin, but more heparin effect. We need to find a way to make sure the patients are anticoagulated.
And ACT, as a test, isn’t necessarily bad. We all know that we don’t use a pump sucker until the ACT is 300, we don’t go on pump until it is 400, and we all have our rules. And that works pretty good.
But the difference is those patients are very consistent, they are usually not sick, they are not hemodiluted, and they have normal AT3 levels. But, in fact, the patients we are putting on ECMO are not normal. Their platelet counts tend to be very low. And thrombocytopenia, in particular, is associated with elevation of the ACT. And what the bedside techs tend to do is start coming down on the heparin. And so it may not be the heparin dose.
And what I think we need to do is look more closely at what is the ACT effect and pay attention to: Do we have a good heparin effect and is the ACT accurate in these circumstances? So that if your patients are septic or have profound thrombocytopenia, you may need to crank the heparin up to counter that.
The other thing that we have noted in the ECMO community, through ELSO, is that when the ACT is difficult to manage, or the heparin is difficult to manage, is that we need to do two other things: We need to reduce source emboli, and we need to reduce situations that can have clot. In other words, not have low-flow states, not turn the pump down.
And the second we have identified is the bridge. Now, the bridge was traditionally put in because everyone was put on VA. And in VV we recognize we don’t need a bridge. And so now what a lot of centers have done is, we do see clots in the bridge, they have either modified their bridge or a lot of us now have taken the bridge out of the circuit until just before weaning, because it clearly is a source of emboli. Which goes along with your data showing that patients who are on longer tend to have more problems, more death and, on autopsy, I suspect tend to have more emboli. And the bridge may be a source. And those of us now who have taken the bridge out are not seeing these emboli in the fingers and toes in the kids who require longer pump runs.
The two questions I have for you are: I am assuming they did, but did all these patients have a bridge, and have you considered taking that out?
And the second is, have you correlated other explanations for coagulation problems, such as platelet count? If a lot of your patients were very thrombocytopenic and the ACTs were high but the heparin was low, that may be an explanation that you did not have a heparin effect, even though the ACT was high, which kind of explains why the ACT may be unreliable.
DR BAIRD: There were bridges in most of all the patients as, I guess, early on in the ECMO experience bridges were there. So most of these patients did have the bridge. And in fact, looking at the autopsy data of about 75 of these patients, I think more than 45 of them had evidence of thromboemboli or embolus. So it is true that maybe taking the bridge out is one way to address the problem in terms of emboli as potentially the cause of death.
Secondly, certainly within this study of 600 patients, being retrospective, looking at things such as platelets would be a little bit difficult for us because platelets aren’t necessarily measured as frequently as ACTs were, which may have been measured hourly. So it would have been a little bit hard, and thus we didn’t necessarily correlate lower platelet levels and correlate them to ACT levels. However, clearly, as you have stated, there are a number of variables, such as sepsis and decreased platelets, and on and on, as an explanation for lowered ACT levels.
Which leads in to what you were alluding to, I guess, in terms of optimizing the ACT—and not necessarily for you to walk away from here saying we should get rid of ACT, because clearly that’s been the standard for a long time—but maybe optimizing the ACT and doing some type of modified ACT test or, in fact, using other tests, more direct tests, of heparin measurements, and things like that, and TEGs, to kind of help clarify our level of anticoagulation.
DR JEFFREY M. PEARL (Cincinnati, OH): We have noticed a difference in ACT measurement depending on the method used. There is a low-range and a high-range cartridge, or a tube-based system. We have found that using the high-range cartridge, which is standardly used in the operating room, was not very accurate on patients on ECMO where you are working at low ACT ranges. So the first question is, does it matter or did you notice a difference in what kind of ACT system was used?
The second observation would be, is the increased heparin usage in the survivors simply a reflection potentially of better preoperative, or pre-ECMO state, or better support on ECMO, better liver function? And that is really what we are seeing as why they had increased survival. The third question is if you can just comment, if ACT is not a good predictor, how does this potentially translate into how we measure patients on cardiopulmonary bypass, where we standardly use the ACT and not necessarily the heparin dosage?
DR BAIRD: With regard to the ACT and the low and high doses, we didn’t necessarily break that out. And I guess one of the negatives of the study is, looking over a long period of time like this, the ways that ACTs were measured were changed. So certainly that is potentially a negative and provides a bit of a variability within the study with regard to ACT. So specifically within this large number of patients, that is certainly a potential drawback and a contributor to the variability of ACTs.
With regard to survival and the amount of heparin, it is probably true that in some way it related to the amount of heparin used, if you’d break down the patients and look at the respiratory patients, for potentially a number of reasons, but the respiratory patients in general required a little more heparin, whether it was because they potentially were, in general, maybe a little bit healthier, in a sense, or also that the cardiac patients maybe had a predisposition to bleeding. But nonetheless, it probably contributes somewhat that the patients that were in a little bit better state of health required differing amounts of heparin.
With regards to ACT and cardiopulmonary bypass, there have been some studies out there suggesting that ACTs—and I think Dr Hines had mentioned this even prior to the talk—that as one goes on cardiopulmonary bypass, you are in a relatively stable state, in other words, maybe even a little bit healthier. So the ACT measurements initially going on bypass may be a little bit more realistic and less variable, as opposed to coming off bypass when a number of different mechanisms have been incited. And so, yes, the ACTs may not be very predictive coming off bypass; however, there also have been studies showing that the higher level the ACT is, maybe the more predictive it is. And we are talking about, here, lower levels of ACTs, in the 180 to 200 range.
And the other thing we are focusing on are pediatric patients, as opposed to adult patients, which I think clearly there has been a difference shown as well in terms of the ACTs.
So in summary, I guess, ultimately, what we’d like to say here is not to go away to change the methods necessarily, but maybe to optimize the methods. And if you look at the amount of heparin you use, clearly you are going to be limited by the state of anticoagulation and the amount of bleeding, but maybe to try to maximize the amount of heparin you use within the range of clinical relevancy.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
