Volume 29, Issue 6 , Pages 616-624, June 2010
Post-operative heparin may not be required for transitioning patients with a HeartMate II left ventricular assist system to long-term warfarin therapy
Article Outline
Background
Anti-coagulation with heparin is often used after left ventricular assist device implantation as a transition to long-term warfarin therapy. We retrospectively evaluated the effects of heparin use on thromboembolic and bleeding complications after implantation of the HeartMate II left ventricular assist device (LVAD).
Methods
LVAD patients (n = 418) implanted as a bridge to transplant were divided into three groups: Group A patients (therapeutic, n = 118) received heparin and had a partial thromboplastin time (PTT) of >50 seconds on two or more occasions; Group B patients (sub-therapeutic, n = 178) had at least one PTT value in the range of 40 to 55 seconds; and Group C patients (no heparin, n = 122) had no PTT values >40 seconds. All patients were transitioned to warfarin and aspirin therapy. The following adverse events were evaluated: ischemic stroke; hemorrhagic stroke; pump thrombosis; bleeding requiring surgery; and bleeding requiring ≥2 units of packed red blood cells in 24 hours.
Results
There was no difference in the percentages of patients with ischemic (5%, 4%, 3%) or hemorrhagic (3%, 3%, 5%) strokes or pump thrombosis (3%, 2%, 2%) after post-operative day (POD) 3 among Groups A, B and C, respectively. From PODs 3 to 30, the percentage of patients requiring transfusion for bleeding was significantly lower for Group C (18%) than for Groups A (32%) and B (26%) (p = 0.04); differences after 30 days were not significant. Multivariate analysis revealed that post-operative heparin use, low post-operative platelet count and low baseline hematocrit value were independent risk factors for bleeding events between PODs 3 and 30.
Conclusions
In patients receiving the HeartMate II LVAD who were directly transitioned to warfarin and aspirin therapy without intravenous heparin there was no short-term increase in risk of thrombotic or thromboembolic events, and bleeding requiring transfusion was significantly reduced. Additional long-term follow-up is needed to evaluate possible late effects.
Keywords: HeartMate II, LVAD, heparin, anticoagulation management, thrombosis
Mechanical circulatory support with a left ventricular assist device (LVAD) is effective therapy for patients with advanced-stage heart failure. Clinical trials with pulsatile LVADs have shown that there is a survival and quality-of-life advantage for LVAD support as compared with continued medical therapy in patients who have a limited life expectancy.1, 2 However, adverse events during support limit the effectiveness and adoption of this therapy. Ongoing research is targeted at refining LVAD technology and patient management protocols to reduce the frequency and severity of adverse events associated with implantable LVADs.
Bleeding and thromboembolism are serious adverse events that have been associated with the use of LVADs in patients with advanced heart failure. Bleeding is the most frequent of these complications and can contribute to the development of other serious adverse events, such as infection and right heart failure. Multiple device- and patient-related factors contribute to the development of post-operative bleeding. The size of the device, the thoracic and abdominal surgery, the use of cardiopulmonary bypass and the requirement for anti-thrombotic therapy throughout support predispose all patients to bleeding complications. Furthermore, LVAD patients often have coagulopathies due to multiple-organ dysfunction and malnutrition, which can result from chronic low cardiac output.
The use of smaller devices with rotary blood pump technology can potentially reduce bleeding by requiring less surgery for implantation. The smaller blood-contacting surface areas of these pumps may also result in less risk of thrombosis. However, the high shear environment increases the potential for platelet activation. The HeartMate II continuous-flow LVAD (Thoratec Corporation, Pleasanton, CA) was developed as a more reliable, long-term support system, substantially smaller than its pulsatile-flow LVAD predecessor, and initial clinical results have been published.3, 4, 5 Because of an initial concern about potential thrombosis, anti-thrombotic therapy was recommended in the treatment protocol and included the early post-operative use of intravenous heparin as a transition to warfarin and aspirin therapy. However, the results of the HeartMate II BTT clinical trial demonstrated that patients had a low risk of thromboembolism,6, 7 whereas post-operative bleeding was the most frequently associated adverse event in the U.S. clinical trial,4, 5 and in the European experience.8, 9 Consequently, physicians began to decrease the amount of anti-coagulant therapy they prescribed.
This study was undertaken to assess the clinical risks and benefits of heparin use by evaluating differences in bleeding and thrombotic events between patients supported by the HeartMate II who did and who did not receive intravenous heparin as a transition to warfarin and aspirin therapy.
Methods
In this retrospective, non-controlled study, data were obtained from 35 centers with a total of 418 patients who had the HeartMate II LVAD implanted as a bridge to transplant (BTT) between March 2005 and January 2008. The HeartMate II LVAD has been described previously.3, 4, 5, 10, 11 All patients were enrolled in the BTT clinical trial sponsored by Thoratec Corporation and approved by the U.S. Food and Drug Administration. Each patient or their representative provided informed written consent before LVAD implantation. All patients in this study had been on LVAD support for >7 days.
Anti-coagulation guidelines
The anti-coagulation guidelines at the beginning of the clinical trial recommended initiation of intravenous heparin on the first post-operative day (POD) or when chest-tube drainage was <50 ml/h. After initiation, the heparin was titrated to achieve a partial thromboplastin time (PTT) of 45 to 50 seconds for 24 hours. In the second 24-hour period (typically PODs 2 to 3), the dose of heparin was titrated upward to achieve a PTT in the range of 50 to 60 seconds. The dose was then continually increased until the PTT reached 55 to 65 seconds. Warfarin was begun when the patient could take oral medications. When the international normalized ratio (INR) was in the range of 2.0 to 3.0, heparin was stopped. On PODs 2 or 3, anti-platelet therapy was begun, consisting of aspirin (81 to 325 mg/day) and dipyridamole (75 mg three times a day). Warfarin, aspirin and dipyridamole were continued throughout support.
The anti-coagulation guidelines were intended to be flexible and could be modified by the investigators to meet specific patient needs or local institution management protocols. Approximately 70% of patients received heparin as a transition to warfarin and aspirin, which was then used throughout the duration of support. Dipyridamole was sometimes prescribed, and a small number of patients received clopidogrel or pentoxifylline. Many of the investigators decreased the amount of anti-coagulation they prescribed over the course of the trial.
Assignment to groups
The PTT and INR were measured on PODs 1, 3, 5, 7, 10, 14 and 21, and at monthly intervals thereafter. Patients were assigned to one of three groups based on PTT values on PODs 3, 5, 7, 10 and 14, defining the therapeutic level of heparin: Group A (therapeutic, n = 118) patients received intravenous heparin and had two or more PTT values >55 seconds; Group B (sub-therapeutic, n = 178) patients received intravenous heparin at a low dose or intermittently and had at least one PTT value >40 seconds, but did not have more than one value >55 seconds; and Group C (no heparin, n = 122) patients did not receive heparin therapy and none had a PTT value >40 seconds. All patients were transitioned to warfarin and aspirin for long-term therapy throughout circulatory support.
Adverse events
The following adverse events were determined for each group: ischemic stroke; hemorrhagic stroke; pump thrombosis; bleeding requiring surgery; and bleeding not requiring surgery. Stroke was defined as a neurologic event lasting >24 hours; the definition of ischemic or hemorrhagic was assigned based on findings from standard imaging studies. Pump thrombosis was defined as thrombus within the device with clinical signs of decreased pump support requiring intervention. Bleeding was defined as an event requiring surgery or requiring ≥2 units of packed red blood cells transfused in a 24-hour period. Adverse events were determined for three time periods: (a) peri-operative events on PODs 0 to 2; (b) during PODs 3 to 30; and (c) after PODs 30. The effects of heparin were determined on the adverse events from PODs 3 to 30 and after POD 30. All adverse events were adjudicated by an independent clinical events committee associated with the BTT clinical trial.
Patient outcomes were determined as survival to transplantation, recovery of ventricular function allowing explant of the device, ongoing device support at 180 days, death or withdrawal from the study.
Statistical analyses
Differences between measures of continuous variables across multiple groups were analyzed using the analysis of variance (ANOVA) F-test if the measures were normally distributed. Otherwise, the non-parametric Kruskal–Wallis test was used instead. Differences between categorical variables were evaluated using Pearson's chi-square test. Univariate analysis was performed with bleeding and thrombotic events (ischemic stroke or pump thrombosis) on PODs 3 to 30 as response variables. Continuous variables were dichotomized at the median value if p < 0.1. Univariate logistic regression was performed on all categorical variables to identify the potential risk factors for early bleeding followed by stepwise forward multivariate logistic regression on the univariate predictors, with an entry criterion of p < 0.1. A multivariate analysis was not performed for thrombotic events due to the small number of events. Concurrent laboratory values (INR, platelet count [PLT] and hematocrit [HCT]) were tracked for each patient post-operatively. For patients with adverse events, the laboratory value at the time of the event was retained, whereas for those without adverse events the average laboratory value from PODs 3 through 21 was used. Statistical significance was set at p < 0.05. All statistical analyses were done using SYSTAT (Cranes Software, Chicago, IL).
Results
Between March 2005 and January 2008, 418 patients at 35 centers had a HeartMate II LVAS implanted as a bridge to transplant. There were 321 men (77%) and 97 women (23%), with an average age of 51.4 ± 13.3 years. The average duration of LVAS support was 293 ± 263 days. Baseline demographic and laboratory data are presented by group in Table 1. Most of the patients in Group C were implanted in the latter half of the trial (after October 2006), and were predominantly from high-volume centers (defined as centers with >15 implants). There were no differences in most baseline values, including platelet count, hematocrit, sodium, creatinine, alanine aminotransferase (ALT) or model end-stage liver disease (MELD) score among the three groups. However, baseline PTT and INR values for Group C were significantly lower than for Groups A and B (p < 0.001), but there were no differences between groups by POD 1 (Figure 1). Post-operatively, the mean PTT values for Group A reached the target therapeutic range on POD 3 and remained in this range through Day 14. Mean PTT values in Group B were elevated, but still sub-therapeutic. Mean PTT values for Group A remained in the normal range for the for the entire 3-week period. All three groups responded to warfarin therapy similarly throughout the post-operative period (PODs 0 to 21) (Figure 2). There was a trend for lower INR at each time-point for Group C compared with Groups A or B. Overall, Group C patients received less anti-coagulant therapy in the entire post-operative period.
Table 1. Baseline Demographics and Laboratory Values for the Three Study Groups
| Characteristic | Group A: therapeutic heparin (n = 118) | Group B: sub-therapeutic heparin (n = 178) | Group C: no heparine (n = 122) | p-value |
|---|---|---|---|---|
| Age (years) | 51 ±12 | 51±13 | 52±15 | 0.40 |
| Women | 25% | 26% | 17% | 0.18 |
| Ischemic etiology | 37% | 44% | 52% | 0.08 |
| Platelet count (1,000s/mm3) | 233±90 | 224±91 | 222±85 | 0.40 |
| Partial thromboplastin time (s) | 53±30 | 52±31 | 43±29 | <0.001 |
| International normalized ratio | 1.40±0.4 | 1.36±0.5 | 1.22±0.2 | <0.001 |
| Implants post-2006(n = 208) | 48(41%) | 85(48%) | 75(61%) | 0.004 |
| Volume >15 implants (n = 203) | 45(38%) | 75(42%) | 83(68%) | <0.001 |
| Hematocrit (%) | 34.5±5.8 | 35.0±5.0 | 34.5±5.6 | 0.4 |
| Sodium (mEq/liter) | 133.4±4.9 | 133.9±5.0 | 133.8±5.0 | 0.8 |
| Creatinine (mg/dl) | 1.5±0.6 | 1.4±0.5 | 1.4±0.4 | 0.5 |
| ALT (U/L) | 81±194 | 124±318 | 71±154 | 0.4 |
| MELD score | 12.76±6.36 | 12.63±6.05 | 11.05±4.89 | 0.07 |
Figure 1.
Mean (± SD) partial thromboplastin time (PTT) values for the three study groups for the first 21 VAD post-operative days. Group A: therapeutic heparin; Group B: sub-therapeutic heparin; Group C: no heparin. POD, post-operative day.
Figure 2.
Mean (± SD) international normalized ratio (INR) values for the three study groups for the first 21 VAD post-operative days. Group A: therapeutic heparin; Group B: sub-therapeutic heparin; Group C: no heparin. POD, post-operative day.
Adverse events for the three groups during the three time periods are listed in Table 2a, Table 2b, Table 2c. There were no significant differences between ischemic and hemorrhagic stroke, pump thrombosis or bleeding requiring surgery in any of the time periods. However, the percentage of patients requiring transfusion for bleeding from PODs 3 to 30 was significantly lower for Group C (18%) than for Groups B (26%) and A (32%) (p = 0.04) (Table 2b). In addition, the total percentages of patients with hemorrhagic events (bleeding requiring surgery, bleeding requiring transfusion, hemorrhagic stroke) were also significantly lower for Group C compared with Groups A and B, and there were no differences in total thrombotic events (pump thrombosis plus ischemic stroke) (Figure 3). There was a trend for less bleeding after 30 days (Table 2c) in Group C, although this was not significant (p = 0.10). When combining all events from PODs 3 to 30 and after POD 30, there were no differences among patients with thrombotic events: ischemic stroke was 5%, 4% and 3% and pump thrombosis was 3%, 2% and 2% for Groups A, B, and C, respectively. In addition, there were no differences in hemorrhagic stroke between groups (3%, 3%, 5%). Bleeding requiring surgery was also not statistically significant (p = 0.40), but there was a trend toward fewer events for Group B (8%) and Group C (10%) when compared with Group A (14%).
Table 2a. Peri-operative Adverse Events for the Three Study Groups for PODs 0 Through 2
| Peri-operative events | Group A: therapeutic heparin (n = 118) | Group B: sub-therapeutic heparin (n = 178) | Group C: no heparin (n = 122) | p-value |
|---|---|---|---|---|
| Ischemic stroke | 2(2%) | 4(2%) | 1(1%) | 0.64 |
| Number of events | 2 | 4 | 1 | |
| Hemorrhagic stroke | 0(0%) | 0(0%) | 0(0%) | N/A |
| Number of events | 0 | 0 | 0 | |
| Pump thrombosis | 0(0%) | 0(0%) | 0(0%) | N/A |
| Number of events | 0 | 0 | 0 | |
| Bleeding requiring surgery | 19(16%) | 17(10%) | 19(16%) | 0.17 |
| Number of events | 21 | 18 | 19 | |
| Bleeding requiring transfusiona | 45(38%) | 59(33%) | 35(29%) | 0.30 |
| Number of events | 47 | 59 | 37 |
aTwo or more of packed red cells in a 24-hour period. |
Table 2b. Adverse Events for the Three Study Groups for PODs 3 to 30
| Peri-operative events | Group A: therapeutic heparin (n = 118)(8.8 pt-years) | Group B: sub-therapeutic heparin (n = 178)(13.1 pt-years) | Group C: no heparin (n = 122)(9.0 pt-years) | p-value |
|---|---|---|---|---|
| Ischemic stroke | 1(0.8%) | 4(2.3%) | 0(0.0%) | 0.20 |
| Events/pt-year | 0.11 | 0.31 | 0 | |
| Hemorrhagic stroke | 2(1.7%) | 2(1.1%) | 1(0.8%) | 0.80 |
| Events/Pt-year | 0.23 | 0.15 | 0.11 | |
| Pump thrombosis | 1(0.9%) | 2(1.1%) | 1(0.8%) | 1.00 |
| Events/pt-year | 0.11 | 0.15 | 0.11 | |
| Bleeding requiring surgery | 11(9.3%) | 10(5.6%) | 9(7.4%) | 0.50 |
| Events/pt-year | 1.37 | 0.76 | 1.00 | |
| Bleeding requiring transfusiona | 38(32%) | 47(26%) | 22(18%) | 0.04 |
| Events/pt-year | 7.66 | 5.89 | 3.99 |
aTwo or more of packed red cells in a 24-hour period. |
Table 2c. Adverse Events for the Three Study Groups After POD 30
| Peri-operative events | Group A: therapeutic heparin (n = 118)(91.7 pt-years) | Group B: sub-therapeutic heparin (n = 178)(112.8 pt-years) | Group C: no heparin (n = 122)(97.4 pt-years) | p-value |
|---|---|---|---|---|
| Ischemic stroke | 5(4.2%) | 3(1.7%) | 4(3.3%) | 0.40 |
| Events/pt-year | 0.07 | 0.03 | 0.05 | |
| Hemorrhagic stroke | 2(1.7%) | 4(2.3%) | 5(4.1%) | 0.47 |
| Events/pt-year | 0.03 | 0.04 | 0.06 | |
| Pump thrombosis | 2(1.7%) | 2(1.1%) | 1(0.8%) | 0.82 |
| Events/pt-year | 0.02 | 0.01 | 0.01 | |
| Bleeding requiring surgery | 6(5.1%) | 5(2.8%) | 3(2.5%) | 0.5 |
| Events/pt-year | 0.07 | 0.04 | 0.03 | |
| Bleeding requiring transfusiona | 34(29%) | 35(20%) | 23(19%) | 0.1 |
| Events/pt-year | 0.80 | 0.74 | 0.47 |
aTwo or more of packed red blood cells in a 24-hour period. |
Figure 3.
Percentage of patients with any bleeding or thrombotic adverse event during the first 3 to 30 days after the operation. Bleeding events = bleeding requiring surgery, bleeding requiring >2 units of packed red blood cells per 24 hours or hemorrhagic stroke; thrombotic events = ischemic stroke or pump thrombosis. Group A: therapeutic heparin; Group B: sub-therapeutic heparin; Group C: no heparin. The p-value indicates the difference between groups according to the chi-square test.
Outcomes for the three study groups are listed in Table 3. There were no significant differences in successful study outcomes of transplantation, recovery or ongoing device support in Groups A (84%), B (86%) or C (92%) (p = 0.08). The causes of death related to thrombotic and hemorrhagic events are shown in Table 4. The number of patients who died of hemorrhagic stroke was the same for all three groups (2 patients per group). The only patient who died of pump thrombosis was receiving therapeutic levels of heparin.
Table 3. Outcomes for All Patients in the Three Study Groups From Implant Through 180 Days
| Characteristic | Group A: therapeutic heparin (n = 118) | Group B: sub-therapeutic heparin (n = 178) | Group C: no heparin (n = 122) | p-value |
|---|---|---|---|---|
| Successful outcomes: transplanted, recovered or ongoing support at 180 days | 99(84%) | 153(86%) | 112(92%) | 0.08 |
| Transplanted | 36(31%) | 65(37%) | 33(27%) | 0.20 |
| Ongoing support | 62(53%) | 86(49%) | 79(65%) | 0.02 |
| Explanted | 1(1%) | 1(0.6%) | 0(0%) | 0.60 |
| Unsuccessful outcomes | 19(16%) | 26(14%) | 10(8%) | 0.15 |
| Died | 19(16%) | 24(13%) | 10(8%) | 0.17 |
| Withdrawn | 0 | 2(0.6%) | 0(0%) | NA |
Table 4. Causes of Death for the Three Study Groups From Implant Through 180 Days
| Cause of death | Group A: therapeutic heparin (n = 118) | Group B: sub-therapeutic heparin (n = 178) | Group C: no heparin (n = 122) |
|---|---|---|---|
| Hemorrhagic stroke | 2(1.7%) | 2(1.1%) | 2(1.6%) |
| Ischemic stroke | 0(0.0%) | 3(1.7%) | 2(1.6%) |
| Pump thrombosis | 1(0.8%) | 0(0.0%) | 0(0.0%) |
| Other causes | 16(13.6%) | 19(10.7%) | 6(4.9%) |
| Total | 19(16%) | 24(13%) | 10(8%) |
Results of the univariate and multivariate analysis of risk factors for bleeding and thrombotic events are depicted in Table 5, Table 6. Low baseline hematocrit (below the median of 34), any heparin use (therapeutic or sub-therapeutic, e.g., Groups A or B) and low platelet count (≤241) were statistically significant independent multivariate risk factors related to bleeding. Area under the receiver operating characteristic curve was 0.63. Heparin use exacerbated the risk of bleeding in patients with low platelet count (any heparin: 36%; no heparin: 20%; p = 0.016), whereas heparin's impact on bleeding was minimal in patients with high platelet count (any heparin: 20%; no heparin: 16%; p = 0.69). Results of risk factors for thrombotic events are summarized in Table 6. Because the number of thrombotic events was small (n = 9), the sample size was not sufficient to perform a multivariate logistic regression. Almost all thrombotic events (PODs 3 to 30) occurred in patients implanted before October 2006, and at lower volume centers. The use of heparin or lack of heparin did not have any impact on thrombotic events.
Table 5. Univariate and Multivariate Analyses of Bleeding Events from PODs 3 to 30a
| Characteristic | Bleeding PODs 3–30 | p-value (univariate) | Odds ratios (univariate) | Odds ratios (multivariate) | p-value (multivariate) | |
|---|---|---|---|---|---|---|
| Yes (n = 107) | No (n = 311) | |||||
| Baseline values | ||||||
| Age | 53±13 | 51±13 | 0.138 | — | — | — |
| Female (n = 97) | 32(33%) | 65(67%) | 0.058 | 1.62[0.98–2.65] | — | — |
| Male (n = 321) | 75(23%) | 246(77%) | ||||
| Ischemic etiology (n = 185) | 43(23%) | 142(77%) | 0.326 | — | — | — |
| Non-ischemic (n = 233) | 64(27%) | 169(73%) | ||||
| PLT (1,000s/mm3) | 214±78 | 230±92 | 0.118 | — | — | — |
| PTT (s) | 51±34 | 49±29 | 0.906 | — | — | — |
| INR | 1.31±0.29 | 1.34±0.44 | 0.923 | — | — | — |
| HCT | 33.4±5.7 | 35.1±5.4 | 0.004 | — | — | — |
| HCT >34(n = 198) | 40(20%) | 158(80%) | 0.015 | 1.75[1.12–2.75] | 1.83[1.15-2.91] | 0.011 |
| HCT ≤34(n = 218) | 67(31%) | 151(69%) | ||||
| Na (mmol/liter) | 134.5±5.2 | 133.5±4.8 | 0.087 | — | — | — |
| Na >134(n = 186) | 56(30%) | 130(70%) | 0.066 | 1.51[0.97–2.35] | — | — |
| Na ≤134(n = 230) | 51(22%) | 179(78%) | ||||
| Creatinine (mg/dl) | 1.51±0.57 | 1.39±0.49 | 0.058 | — | — | — |
| Creatinine >1.30(n = 199) | 59(30%) | 140(70%) | 0.080 | 1.48[0.95–2.31] | — | — |
| Creatinine ≤1.30(n = 265) | 48(22%) | 217(78%) | ||||
| ALT (IU) | 78±190 | 103±264 | 0.261 | — | — | — |
| Implants after October 2006 (n = 210) | 45(22%) | 78% | 0.065 | 1.52[1.97–2.38] | — | — |
| Implants before October 2006 (n = 208) | 62(30%) | 70% | ||||
| Volume >15 implants (n = 203) | 46(23%) | 157(77%) | 0.182 | 1.35[0.87–2.08] | — | — |
| Volume ≤15 implants (n = 205) | 61(28%) | 154(72%) | ||||
| MELD | 12.5±5.9 | 12.1±5.8 | 0.342 | — | — | — |
| On LVAD support | ||||||
| Peri-operative bleeding (n = 139)b | 44(32%) | 95(68%) | 0.046 | 1.59[1.00–2.50] | ||
| No peri-operative bleeding (n = 279) | 63(23%) | 216(77%) | ||||
| Any heparin (Groups A + B) | 85(29%) | 211(71%) | 0.024 | 1.83[1.08–3.10] | 1.80[1.05–3.08] | 0.033 |
| No heparin (Group C) | 22(18%) | 100(82%) | ||||
| INRc | 1.78±0.99 | 1.68±0.49 | 0.131 | — | — | — |
| PLTc | 240±149 | 256±92 | 0.002 | — | — | — |
| PLTc >241 | 39(19%) | 166(81%) | 0.005 | 1.92[1.22–3.03] | 1.94[1.22–3.09] | 0.005 |
| PLTc ≤241 | 65(31%) | 144(69%) | ||||
| HCTc | 28.8±4.0 | 29.8±3.5 | 0.003 | — | — | — |
| HCTc >29.2 | 39(19%) | 168(81%) | 0.003 | 1.97[0.80–3.11] | — | — |
| HCTc ≤29.2 | 65(31%) | 142(69%) | ||||
aRisk factors with p < 0.1 were used in the multivariate model. |
bPeri-operative bleeding on PODs 0–2 prior to the study period of PODs 3–30. |
cLaboratory value at the time of the event for patients with bleeding events, and the average value from PODs 3 through 21 for patients without bleeding events. |
Table 6. Univariate Analysis of Thrombotic Events During PODs 3 to 30a
| Characteristic | Thrombotic events at PODs 3–30 | p | |
|---|---|---|---|
| Yes (n = 9) | No (n = 409) | ||
| Baseline parameters | |||
| Age | 46±17 | 52±13 | 0.242 |
| Female (n = 97) | 4(4%) | 93(96%) | 0.142 |
| Male (n = 321) | 5(2%) | 316(98%) | |
| Ischemic etiology (n = 185) | 4(2%) | 181(98%) | 0.991 |
| Non-ischemic (n = 233) | 5(2%) | 228(98%) | |
| PLT (1,000s/mm3) | 240±53 | 226±89 | 0.639 |
| PTT (s) | 40.3±13.4 | 49.8±30.8 | 0.279 |
| INR | 1.88±1.64 | 1.32±034 | 0.018 |
| HCT | 38.0±5.3 | 34.6±5.5 | 0.084 |
| Na | 134±4 | 134±5 | 0.700 |
| Creatinine (mg/dl)a | 1.21±0.44 | 1.42±0.51 | 0.184 |
| ALT | 305±775 | 92±222 | 0.082 |
| Implants after October 2006(n = 210) | 1(0.5%) | 209(99.5%) | 0.046 |
| Implants before October 2006(n = 208) | 8(4%) | 200(96%) | |
| Volume >15 implants (n = 203) | 0(0%) | 203(100%) | 0.004 |
| Volume ≤15 implants (n = 205) | 9(4%) | 206(96%) | |
| MELD | 13.5±10.8 | 12.2±5.7 | 0.532 |
| During LVAD support | |||
| No heparin (n = 122) | 1(1%) | 121(99%) | 0.256 |
| Any heparin (n = 296) | 8(3%) | 288(97%) | |
| PTTb | 45.4±123.6 | 44.3±13.7 | 0.805 |
| INRb | 2.28±0.79 | 1.68±0.50 | 0.015 |
| PLTb | 245±119 | 248±91 | 0.845 |
| HCTb | 29.9±4.8 | 29.5±3.5 | 0.756 |
aDue to the small number of patients with a thrombotic event, a multivariate analysis was not performed. |
bLaboratory value at the time of the event for patients with bleeding events, and the average value from PODs 3 through 21 for patients without bleeding events. |
Discussion
The results of this study indicate that patients who do not receive early post-operative anti-coagulation therapy with intravenous heparin as a transition to warfarin and aspirin therapy are not at any early increased risk for thrombotic events, and their risk of post-operative bleeding is reduced. Patients who receive intravenous heparin in transition to warfarin therapy have the same rate of thrombotic events, but are at increased risk of post-operative bleeding requiring transfusion based on our findings. The effects of heparin on bleeding are exacerbated in patients with low post-operative platelet count. Therefore, eliminating the routine use of post-operative heparin in patients with a low risk of thrombosis appears to be appropriate for most cases.
At the beginning of the U.S. HeartMate II clinical trial, it was assumed that the device needed a rigorous anti-coagulation protocol. However, studies have since shown a low risk of thrombotic events during HeartMate II support.4, 5, 6, 7 In a recent analysis of 281 patients supported by the HeartMate II, ischemic stroke occurred in 15 patients (5%), and there were 4 instances (1%) of primary device thrombosis.4 A separate, independent clinical trial in Europe yielded similar results.8, 9 A recent analysis of outpatient anti-coagulation in these patients also concluded that the small number of thrombotic events was overshadowed by a larger number of bleeding events.7
Although the recent clinical results indicate that there is a low rate of thrombotic events during HeartMate II support, bleeding was the most frequent adverse event in both the U.S. and European studies, occurring in about 50% of all cases. These results are consistent with data from a worldwide registry, which includes a variety of ventricular assist devices.12 Bleeding also contributes to the development of other severe complications, such as infection and multiple-organ failure.
Post-operative bleeding is challenging to manage because multiple factors contribute to alterations in the coagulation system, which can fluctuate between hypo- and hypercoagulability after VAD implantation. By the time patients receive their LVAD, chronic heart failure has often affected other end-organs, such as the kidney and liver, which can lead to the altered coagulability that often occurs after implant.13 Intra-operatively, the coagulation system is affected by platelet activation, thrombin generation, fibrinogen cleavage and activation of fibrinolysis.14 In the immediate post-operative period, patients may develop thrombocytopenia, platelet dysfunction and fibrinolysis. Thus, both intra- and post-operatively, patients are at increased risk for hypocoagulability. Although it appears that there is a gradual recovery toward normal coagulation after the implantation,15 there is evidence that the LVAD may impair platelet function and can also induce acquired von Willebrand syndrome, which contributes to a constant state of hypocoagulability.16, 17
Although the post-implant hypocoagulable state contributes to bleeding complications, it may protect against thrombus formation. Because of this low risk of thrombosis and higher rate of bleeding events, investigators in the HeartMate II BTT trial began to decrease anti-coagulant and/or anti-platelet therapy as the study period progressed. Most centers have now discontinued anti-platelet therapy with dipyridamole during LVAD support. In a recent analysis of 331 discharged patients supported by the HeartMate II, the number of bleeding events (58) exceeded thrombotic events,10 leading to a recommended decrease in the INR range from 2.0 to 3.0 down to 1.5 to 2.5.7
In patients who have a higher risk of post-implant thrombotic events, for example, patients with atrial fibrillation or artificial heart valves, heparin therapy as a transition to warfarin should still be considered, as should increased levels of anti-coagulation during pump support. Patients who have low VAD flows are also at higher risk of developing pump thrombosis, and increased anti-coagulation during these periods is also recommended.
This study was limited by its retrospective, non-controlled design, which may have resulted in a selection bias. The use of multivariate logistic regression minimizes some of this bias by controlling for potential confounding variables, but there may still be some differences that may not have been captured in the analysis. The frequency of use and dosages of heparin varied and could only be assessed by the PTT values determined at intervals defined by the study protocol and at the time of the event. There is no guarantee that these values are representative of the hematologic state between the protocol intervals. In addition, the anti-coagulation protocol recommended for the study was not consistently followed and there was a variable use of anti-platelet medications. The use of anti-coagulation guidelines rather than a firm protocol provided flexibility for the investigators to meet various patient needs, and the accrued clinical experience changed the rationale for therapy over time.
In conclusion, patients supported by the HeartMate II who transitioned to oral warfarin and aspirin therapy without intravenous heparin did not have an increased risk of thrombotic events and they had significantly fewer bleeding events. The temporary use of intravenous heparin during the transition to warfarin may not be necessary in patients with a low thrombotic risk.
Disclosure statement
M.S.S. reports receiving grant support from Thoratec and HeartWare. D.J.F. is an employee and stockholder of Thoratec. K.S.S. is an employee of Thoratec. J.V.C. reports receiving training and consulting from Thoratec. S.D.R. has received training and consulting support from Thoratec. F.D.P. has received training and research support from Thoratec.
References
- Long-term mechanical left ventricular assistance for end-stage heart failure. N Engl J Med. 2001;345:1435–1443
- Multicenter clinical evaluation of the HeartMate vented electric left ventricular assist system in patients awaiting heart transplantation. J Thorac Cardiovasc Surg. 2001;122:1186–1195
- Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
- Extended mechanical circulatory support with a continuous-flow rotary left ventricular assist device. J Am Coll Cardiol. 2009;54:312–321
- Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med. 2009;361:2241–2251
- Low thromboembolic risk for patients with the HeartMate II left ventricular assist device. J Thorac Cardiovasc Surg. 2008;136:1318–1323
- Low thromboembolism and pump thrombosis with the HeartMate II left ventricular assist device: analysis of outpatient anti-coagulation. J Heart Lung Transplant. 2009;28:881–887
- Lahpor J, Khaghani A, Hetzer R, et al. European results with a continuous-flow ventricular assist device for advanced heart-failure patients. Eur J Cardiothorac Surg (in press).
- HeartMate II left ventricular assist device; early European experience. Eur J Cardiothorac Surg. 2008;34:289–294
- HeartMate II left ventricular assist system: from concept to first clinical use. Ann Thorac Surg. 2001;71(suppl):S116–S120
- HeartMate left ventricular assist devices: a multigeneration of implanted blood pumps. Artif Organs. 2001;25:422–426
- INTERMACS database for durable devices for circulatory support: first annual report. J Heart Lung Transplant. 2008;27:1065–1072
- Vitamin K reduces bleeding in left ventricular assist device recipients. J Heart Lung Transplant. 1999;18:346–350
- Increased activation of the coagulation and fibrinolytic systems leads to hemorrhagic complications during left ventricular assist implantation. Circulation. 1996;94:II-227–II-234
- Fibrinolytic activation during long-term support with the HeartMate II left ventricular assist device. ASAIO J. 2008;54:115–119
- Non-surgical bleeding in patients with ventricular assist devices could be explained by acquired von Willebrand disease. Eur J Cardiothorac Surg. 2008;33:679–684
- Severely impaired von Willebrand factor-dependent platelet aggregation in patients with a continuous-flow left ventricular assist device (HeartMate II). J Am Coll Cardiol. 2009;53:2162–2167
PII: S1053-2498(10)00074-4
doi:10.1016/j.healun.2010.02.003
© 2010 International Society for Heart and Lung Transplantation. Published by Elsevier Inc. All rights reserved.
Volume 29, Issue 6 , Pages 616-624, June 2010
