Activated partial thromboplastin time overestimates anti-coagulation in left ventricular assist device patients

Abstract

Patients with continuous-flow left ventricular assist devices (LVADs) require long-term systemic anti-coagulation after implantation to prevent thrombotic events. Thrombosis of an LVAD may result in cerebrovascular accidents, device exchange or death. Over the past several years, there has been a reported increase in the risk of device thrombosis in the United States, without a clear explanation for the increase.1Starling R.C. Moazami N. Silvestry S.C. et al.Unexpected abrupt increase in left ventricular assist device thrombosis.N Engl J Med. 2014; 370: 33-40Crossref PubMed Scopus (604) Google Scholar Proper anti-coagulation management in patients with an LVAD is essential for thrombosis prevention. Unfractionated heparin (UFH) is often used to bridge LVAD patients early after surgical implantation or when oral anti-coagulation is sub-therapeutic. Due to its mechanism of action, the activated partial thromboplastin time (aPTT) reflects the function of heparin’s effects on the intrinsic pathways of the coagulation cascade, whereas the anti-factor Xa (anti-Xa) assay measures heparin’s impact on anti-thrombin.2Vandiver J.W. Vondracek T.G. Antifactor Xa levels versus activated partial thromboplastin time for monitoring unfractionated heparin.Pharmacotherapy. 2012; 32: 546-558Crossref PubMed Scopus (123) Google Scholar The aPTT is currently the most commonly used laboratory test for monitoring UFH. A large number of variables can affect the aPTT, rendering the patient receiving UFH at risk of supra- or sub-therapeutic anti-coagulation. Routine monitoring of UFH using the anti-Xa assay has been reported to provide a more accurate reflection of anti-coagulation.2Vandiver J.W. Vondracek T.G. Antifactor Xa levels versus activated partial thromboplastin time for monitoring unfractionated heparin.Pharmacotherapy. 2012; 32: 546-558Crossref PubMed Scopus (123) Google Scholar Since the 1990s, the American College of Chest Physicians and the College of American Pathologists have recommended that aPTT goals be titrated to a corresponding anti-Xa level according to individual institutions. It is currently unknown whether the aPTT and anti-Xa levels reflect the same level of anti-coagulation in LVAD patients. Due to a higher-than-expected LVAD thrombosis rate at our institution, we hypothesized that the aPTT may not adequately reflect the level of anti-coagulation with UFH in LVAD patients; therefore, we simultaneously measured the aPTT and anti-Xa activity in both LVAD recipients and patients admitted with acute decompensated heart failure (ADHF). We performed a prospective, single-center quality improvement project that included all hospitalized patients receiving UFH who had (1) an LVAD or (2) ADHF on the Advanced Heart Failure Service. Patients were included if they were receiving UFH after device implantation, as bridging therapy for a sub-therapeutic international normalized ratio (INR), or for active thrombosis. Target INR goals for both the HeartMate II (Thoratec Corporation, Pleasanton, CA) and HeartWare HVAD (HeartWare International Inc, Framingham, MA) were 2.0 to 3.0. We obtained simultaneous aPTT and anti-Xa levels after patients were on a stable dose of UFH for a minimum of 6 hours. aPTT and anti-Xa activity levels were determined by standard techniques using Stago reagents. Therapeutic values were locally defined as aPTT of 60 to 90 seconds and anti-Xa of 0.3 to 0.7 U/ml. Each aPTT value and its corresponding anti-Xa level were plotted. The slope representing the relationship between aPTT levels and anti-Xa levels was determined for each population using linear regression models. Analysis of covariance models were used to determine statistical significance between the groups. All analyses were performed using SAS (version 9.3) statistical software (SAS Institute, Inc., Cary, NC). Our study was approved by the institutional review board of the University of Colorado. From February to June 2014, 19 individual patients with durable LVADs and 10 individual patients with ADHF had simultaneous aPTT and anti-Xa levels drawn (Table 1). The ages (mean ± SD) of the groups were 55 ± 11 years and 59 ± 13 years, respectively. Of the patients with durable LVADs, 9 had been implanted <30 days, 15 had a HeartMate II device, and 4 were being treated for an active device thrombosis.Table 1Demographics and Anti-coagulation Levels of LVAD and ADHF CohortsaShaded values = therapeutic aPTT levels (60 to 90 seconds).LVADAnti-XaaPTTINRWarfarinbNo warfarin therapy defined as >7 days from last dose of warfarin.LVAD <30 daysActive thrombosiscThrombosis defined as: LVAD patients = device thrombosis; ADHF patients = venous thrombosis.Age (years)GenderEtiologyLVAD patients1HMII<0.1113.02.3YNN55FNICM2HMII<0.172.91.8YNN63MICM3HMII<0.147.81.2YYN70MICM4HMII0.15113.31.3YYN52MNICM5HMII0.2878.81.3NYY54MICM6HMII<0.134.21.3YYN72MICM7HMII0.2285.11.5YNN65FNICM8HMII0.1139.41.3NYN53MNICM9HMII0.3128.71.7YNN51MICM10HMII0.1971.41.5YNN66MNICM11HVAD<0.176.21.8YYN59FNICM12HVAD0.1870.31.9YNN40MNICM13HMII<0.169.21.9YYN63MICM14HMII0.43112.73.7YNY50MNICM15HVAD0.2764.91.3YYN32MNICM16HMII0.1363.1—YNN63MNICM17HMII0.26130.22.6YNY57FICM18HMII0.23183.94.4YNY47MNICM19HVAD0.2981.61.2YYN32MNICMADHF patients1—0.3170.52Y—N66MNICM2—0.1650.81.1N—N59FNICM3—0.2859.91.2N—N62MNICM4—0.1659.41.3Y—N65MNICM5—0.2485.2—Y—N39MICM6—0.3980.71.2N—N59MICM7—1.1463.5—N—N81MICM8—0.4771.71.1Y—N36FNICM9—0.3497.81.8Y—N60FNICM10—0.3467.01.2N—N63MICMADHF, acute decompensated heart failure; anti-Xa, anti-factor Xa; aPTT, activated partial thromboplastin time; HMII, HeartMate II; HVAD, HeartWare left ventricular assist device; ICM, ischemic cardiomyopathy; INR, international normalized ratio; LVAD, left ventricular assist device; NICM, non-ischemic cardiomyopathy.a Shaded values = therapeutic aPTT levels (60 to 90 seconds).b No warfarin therapy defined as >7 days from last dose of warfarin.c Thrombosis defined as: LVAD patients = device thrombosis; ADHF patients = venous thrombosis. Open table in a new tab ADHF, acute decompensated heart failure; anti-Xa, anti-factor Xa; aPTT, activated partial thromboplastin time; HMII, HeartMate II; HVAD, HeartWare left ventricular assist device; ICM, ischemic cardiomyopathy; INR, international normalized ratio; LVAD, left ventricular assist device; NICM, non-ischemic cardiomyopathy. Of the 19 patients with LVADs who had simultaneous aPTT and anti-Xa levels drawn, 10 had therapeutic aPTTs. All 10 of these patients had sub-therapeutic anti-Xa levels. Nine of the 10 patients with therapeutic aPTT levels were taking warfarin, with a mean INR of 1.6. Of the 6 patients with supra-therapeutic aPTT levels, 4 had sub-therapeutic and 2 had therapeutic anti-Xa levels. Of the 10 patients with ADHF who had simultaneous aPTT and anti-Xa levels drawn, 6 had therapeutic aPTTs. Of these 6 patients, 4 had therapeutic anti-Xa levels, 1 had a sub-therapeutic anti-Xa level and 1 had a supra-therapeutic anti-Xa level. Three of the 6 patients with therapeutic aPTT levels were taking warfarin, with a mean INR of 1.6. Figure 1 compares aPTT to anti-Xa activity for LVAD patients (Figure 1A) and ADHF patients (Figure 1B), and ADHF patients with the removal of one major outlier (Patient 7) (Figure 1C). The outlier patient was an 81-year-old man with ischemic cardiomyopathy, who was being treated for a bacteremia and was managed on concomitant UFH and a glycoprotein IIb/IIIa inhibitor; however, there was no clear cause for the supra-therapeutic anti-Xa level. Comparison between slopes of the line showed: LVAD vs ADHF (with outlier), p = 0.12; and LVAD vs ADHF (without outlier), p = 0.03. Our findings suggest that the aPTT overestimates the level of UFH anti-coagulation in LVAD patients compared with patients without LVADs hospitalized with ADHF. Three possible explanations for this finding include: (1) an effect of acquired von Willebrand’s (VW) syndrome in lowering Factor VIII (FVIII) levels in LVAD patients; (2) under-dosing UFH due to the effect of warfarin on aPTT levels in LVAD patients; and (3) a combination of both. The presence of acquired VW syndrome in LVAD patients has been reported by a number of investigators.3Crow S. Chen D. Milano C. et al.Acquired von Willebrand syndrome in continuous-flow ventricular assist device recipients.Ann Thorac Surg. 2010; 90: 1263-1269Abstract Full Text Full Text PDF PubMed Scopus (275) Google Scholar The VW factor affects fibrin clot formation by acting as a carrier protein for FVIII. An abnormality in VW factor allows for increased proteolysis of FVIII, resulting in a prolonged aPTT, but with no effect on the anti-Xa activity.4Brinkhous K.M. Sandberg H. Garris J.B. et al.Purified human factor VIII procoagulant protein: comparative hemostatic response after infusions into hemophilic and von Willebrand disease dogs.Proc Natl Acad Sci USA. 1985; 82: 8752-8756Crossref PubMed Scopus (132) Google Scholar A defect in VW factor has been reported to occur in virtually all patients with continuous-flow LVADs.3Crow S. Chen D. Milano C. et al.Acquired von Willebrand syndrome in continuous-flow ventricular assist device recipients.Ann Thorac Surg. 2010; 90: 1263-1269Abstract Full Text Full Text PDF PubMed Scopus (275) Google Scholar To our knowledge, FVIII levels in LVAD patients with acquired VW syndrome have not been reported. Warfarin has a small effect on the aPTT, but the effect is magnified in the presence of heparin. It has been estimated that for each increase of 1.0 in the INR, the aPTT increases by 16 seconds.5Kearon C. Johnston M. Moffat K. et al.Effect of warfarin on activated partial thromboplastin time in patients receiving heparin.Arch Intern Med. 1998; : 1140-1143Crossref PubMed Scopus (37) Google Scholar However, this does not seem likely to be the sole explanation for our findings, because, even when correcting for this difference in our sample, 7 of the 10 LVAD patients would still have therapeutic aPTT levels with sub-therapeutic anti-Xa levels. All durable LVAD patients require long-term systemic anti-coagulation with wafarin therapy; therefore, the aPTT may be falsely elevated and may lead to insufficient anti-coagulation if dosing UFH from aPTT levels. This is a preliminary report in a small number of patients. However, the incidence of thrombosis is high in LVAD patients and all potential causes need to be reported quickly so that they can be investigated in larger numbers and at multiple institutions. Several hypotheses need to be tested in LVAD patients, including: exclusive monitoring of anti-Xa levels (or a combination of anti-Xa and aPTT) and the increased risk of bleeding when monitoring anti-Xa levels. Our results suggest that aPTT levels underestimate anti-coagulation in LVAD patients and monitoring UFH with anti-Xa levels more accurately reflects the level of anti-coagulation. The authors have no conflicts of interest to disclose. A.V.A. was supported by a Scientist Development Grant from the American Heart Association and by the Boettcher Foundation’s Webb-Waring Biomedical Research Program.