Theodore E Warkentin

McMaster University, Hamilton, Ontario, Canada

Are you Theodore E Warkentin?

Claim your profile

Publications (308)2340.27 Total impact

  • Theodore E Warkentin
    New England Journal of Medicine 08/2015; 373(7):642-55. DOI:10.1056/NEJMra1316259 · 55.87 Impact Factor
  • I. Nazi · D. M. Arnold · T. E. Warkentin · J. W. Smith · P. Staibano · J. G. Kelton
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Many patients exposed to heparin develop antibodies against platelet factor 4 (PF4) and heparin, yet only those antibodies that activate platelets cause heparin-induced thrombocytopenia (HIT). Patients who produce anti-PF4/heparin antibodies without developing HIT either have antibodies that do not cause platelet activation or produce pathogenic antibodies at levels that are insufficient to cause HIT. Understanding the differences between anti-PF4/heparin antibodies with and without HIT will improve test methods and reduce overdiagnosis.Objectives To investigate the presence of low levels of platelet-activating antibodies in patients investigated for HIT who had anti-PF4/heparin antibodies but failed to cause platelet activation in the 14C-serotonin release assay (SRA).Methods We developed a platelet activation assay similar to the SRA using exogenous PF4 without added heparin (PF4-SRA). This assay was able to detect low levels of platelet-activating antibodies. We used this PF4-SRA to test for platelet-activating antibodies in patients investigated for HIT.ResultsThe PF4-SRA detected platelet-activating antibodies in 7/7 (100%) SRA-positive sera even after the samples were diluted until they were no longer positive in the standard SRA. Platelet-activating antibodies were detected in 14/39 (36%) patients who had anti-PF4/heparin antibodies but tested negative in the SRA and did not have clinical HIT. The clinical diagnosis of HIT was confirmed by chart review and concordant with the SRA results.ConclusionsA subset of heparin-treated patients produce subthreshold levels of platelet-activating anti-PF4/heparin antibodies that do not cause HIT. An increase in the titre of these pathogenic antibodies, along with permissive clinical conditions could lead to HIT.This article is protected by copyright. All rights reserved.
    Journal of Thrombosis and Haemostasis 08/2015; DOI:10.1111/jth.13066 · 5.72 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Methylene blue pathogen inactivation and storage of thawed plasma both lead to changes in the activity of several clotting factors. We investigated how this translates into a global loss of thrombin generation potential and alterations in the protein C pathway. Fifty apheresis plasma samples were thawed and each divided into three subunits. One subunit was stored for 7 days at 4 °C, one was stored for 7 days at 22 °C and one was stored at 4 °C after methylene blue/light treatment. Thrombin generation parameters, ProC(®)Global-NR, prothrombin time and activated partial thromboplastin time were assessed on days 0 and 7. The velocity of thrombin generation increased significantly after methylene blue treatment (increased thrombin generation rate; time to peak decreased) and decreased after storage (decreased thrombin generation rate and peak thrombin; increased lag time and time to peak). The endogenous thrombin generation potential remained stable after methylene blue treatment and storage at 4 °C. Methylene blue treatment and 7 days of storage at 4 °C activated the protein C pathway, whereas storage at room temperature and storage after methylene blue treatment decreased the functional capacity of the protein C pathway. Prothrombin time and activated partial thromboplastin time showed only modest alterations. The global clotting capacity of thawed plasma is maintained at 4 °C for 7 days and directly after methylene blue treatment of thawed plasma. Thrombin generation and ProC(®)Global are useful tools for investigating the impact of pathogen inactivation and storage on the clotting capacity of therapeutic plasma preparations.
    Blood transfusion = Trasfusione del sangue 07/2015; DOI:10.2450/2015.0030-15 · 2.37 Impact Factor
  • Theodore E Warkentin · Richard J Cook · Ravi Sarode · Debi A Sloane · Mark A Crowther
    [Show abstract] [Hide abstract]
    ABSTRACT: Venous limb gangrene can occur in cancer patients, but the clinical picture and pathogenesis remain uncertain. We identified 10 patients with metastatic cancer (7 pathologically-proven) who developed severe venous limb ischemia (phlegmasia/venous limb gangrene) after initiating treatment for DVT; in 8 patients, cancer was not known or suspected at presentation. The patients exhibited a novel, clinically-distinct syndrome: warfarin-associated supratherapeutic INR (median, 6.5) at onset of limb ischemia; rising platelet count during heparin anticoagulation; and platelet fall after stopping heparin. Despite supratherapeutic INRs, patient plasma contained markedly elevated thrombin-antithrombin (TAT) complex levels (indicating uncontrolled thrombin generation) and protein C (PC) depletion; this profile resembles the greatly elevated TAT/PC activity ratios reported in patients with warfarin-associated venous limb gangrene complicating HIT. Analyses of vitamin K-dependent factors in 6 cancer patients with available serial plasma samples showed that variations in the INR corresponded most closely with changes in factor VII, with a highly collinear relationship between VII and PC. We conclude that venous limb ischemia/gangrene is explained in some cancer patients by profoundly disturbed procoagulant-anticoagulant balance, whereby warfarin fails to block cancer-associated hypercoagulability while nonetheless contributing to severe PC depletion, manifest as a characteristic supratherapeutic INR caused by parallel severe factor VII depletion. Copyright © 2015 American Society of Hematology.
    Blood 05/2015; 126(4). DOI:10.1182/blood-2015-01-622787 · 10.45 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Rapid exclusion of heparin-induced thrombocytopenia (HIT) is needed to determine which patients can continue to receive heparin. In this prospective management study, 526 participants had a 4Ts score, rapid immunoassay (PF4/H-PaGIA) and serotonin-release assay (SRA) performed. While awaiting SRA results, participants with low 4Ts score (irrespective of PF4/H-PaGIA result) or intermediate 4Ts score plus a negative PF4/H-PaGIA result received prophylactic doses of danaparoid or fondaparinux; all others received therapeutic doses of nonheparin anticoagulants. The primary outcome was the frequency of management failures defined as HIT positive participants with a low 4Ts score (irrespective of PF4/H-PaGIA result) or intermediate 4Ts score plus negative PF4/H-PaGIA result. Six participants (1.1%; 95%CI,0.2-2.1%) were management failures. A negative PF4/H-PaGIA result reduced the pretest probability of HIT from 1.9% to 0% (95%CI, 0-1.3%), 6.7% to 0% (95%CI, 0-2.7%) and 36.6% to 0% (95%CI, 0-14.3%) in the low, intermediate and high score groups, respectively. A positive PF4/H-PaGIA increased the probability of HIT in the low score group to 15.4% (95%CI, 5.9-30.5). Thus, a low or intermediate 4Ts score plus negative PaGIA excluded HIT, whereas any other combination of results justified use of alternative anticoagulants until HIT could be excluded. This trial was registered at #NCT00489437. Copyright © 2015 American Society of Hematology.
    Blood 04/2015; 126(5). DOI:10.1182/blood-2014-12-618165 · 10.45 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Heparin-induced thrombocytopenia (HIT) is caused by platelet-activating antibodies against complexes of platelet factor 4 (PF4) and heparin. The diagnosis of HIT is contingent on accurate and timely laboratory testing. Recently, alternative anticoagulants for the treatment of HIT have been introduced along with algorithms for better HIT diagnosis. However, the increased reliance on immunoassays for the diagnosis of HIT may have harmful consequences due to the high rate of false positive results. To compare trends and implications of current HIT testing approaches, we analyzed results over a six-year period from the McMaster University Platelet Immunology Reference Laboratory. From 2008 to 2013, 8546 samples were investigated for HIT using both an in-house IgG-specific anti-PF4/heparin enzyme immunoassay (EIA) and the serotonin-release assay (SRA). Of 8546 samples tested, 13.4% were true-positives (positive in both assays); 65.6% were true-negatives (negative in both assays); 20.9% were presumed false positive for HIT (EIA-positive/SRA-negative); and 0.2% were EIA-negative/SRA-positive. The frequency of EIA-positive/SRA-negative results increased over time (from 12.9% in 2008 to 22.9% in 2013). We found that the number of SRA-negative samples was reduced from referring centers that used an immunoassay as an initial screen; however, 41% of those samples tested negative in the immunoassay and in the SRA at the reference laboratory. The suspicion of HIT continues at a high rate and the agreement between the EIA and SRA test results remains problematic. This article is protected by copyright. All rights reserved. © 2015 Wiley Periodicals, Inc.
    American Journal of Hematology 03/2015; 90(7). DOI:10.1002/ajh.24025 · 3.80 Impact Factor
  • Twylla Tassava · Theodore E Warkentin
    American Journal of Hematology 03/2015; 90(8). DOI:10.1002/ajh.24018 · 3.80 Impact Factor
  • Theodore E Warkentin · Donald M Arnold · Ishac Nazi · John G Kelton
    [Show abstract] [Hide abstract]
    ABSTRACT: Few laboratory assays are as clinically useful as the platelet serotonin-release assay (SRA): a positive SRA in the appropriate clinical context is virtually diagnostic of heparin-induced thrombocytopenia (HIT), a life- and limb-threatening prothrombotic disorder caused by anti-platelet factor 4 (PF4)/heparin antibodies that activate platelets, thereby triggering serotonin-release. The SRA's performance characteristics include high sensitivity and specificity, although caveats include indeterminate reaction profiles (observed in ∼4% of test sera) and potential for false-positive reactions. As only a subset of anti-PF4/heparin antibodies detectable by enzyme-immunoassay (EIA) are additionally platelet-activating, the SRA has far greater diagnostic specificity than the EIA. However, requiring a positive EIA, either as an initial screening test or as an SRA adjunct, will reduce risk of a false-positive SRA (since a negative EIA in a patient with a "positive" SRA should prompt critical evaluation of the SRA reaction profile). The SRA also provides useful information on whether a HIT serum produces strong platelet activation even in the absence of heparin: such heparin-"independent" platelet activation is a marker of unusually severe HIT, including delayed-onset HIT and severe HIT complicated by consumptive coagulopathy with risk for microvascular thrombosis. This article is protected by copyright. All rights reserved. © 2015 Wiley Periodicals, Inc.
    American Journal of Hematology 03/2015; 90(6). DOI:10.1002/ajh.24006 · 3.80 Impact Factor
  • Theodore E. Warkentin · Andreas Greinacher · Jürgen Bux
    Transfusion 02/2015; 55(5). DOI:10.1111/trf.12994 · 3.23 Impact Factor
  • Source
    Theodore E Warkentin · Eric L Safyan · Lori-Ann Linkins
    New England Journal of Medicine 01/2015; 372(5):492-4. DOI:10.1056/NEJMc1414161 · 55.87 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Heparin-induced thrombocytopenia (HIT) is a serious complication of heparin utilization. An enzyme-linked immunosorbent assay (ELISA) is usually performed to assist in the diagnosis of HIT. ELISAs tend to be sensitive but lack specificity. We sought to utilize a new cut-off to define a positive HIT ELISA. We conducted a prospective observational study of hospitalized patients undergoing ELISA testing. All patients who underwent ELISA testing were eligible for inclusion (n=496). Irrespective of the results, all subjects had confirmatory testing with a serotonin release assay (SRA). We compared a threshold optical density (OD)>1.00 to the current definition of a positive ELISA (OD>0.40) as a screening test for a positive SRA. We used sensitivity, specificity, and area under the receiver operating curve to determine whether an OD>1.00 would improve diagnostic accuracy for HIT. The SRA was positive in 10 patients (prevalence: 2.0%). Adjusting the definition of a positive HIT ELISA to >1.00 maintained the sensitivity and negative predictive value at 100% in our cohort. The positive predictive value (PPV) of the higher cutoff OD was more than triple the PPV of an OD>0.40 (41.7% vs 13.3%). No patient with a positive SRA had an OD measurement <1.00. Increasing the OD threshold enhances specificity without noticeably compromising sensitivity. Altering the definition of the HIT ELISA could prevent unnecessary testing and/or treatment with non-heparin based anticoagulants in patients with possible HIT. (NCT 00946400).
    Chest 01/2015; 148(1). DOI:10.1378/chest.14-1417 · 7.48 Impact Factor
  • Theodore E. Warkentin
    Thrombosis and Haemostasis 01/2015; 113(3). DOI:10.1160/TH14-11-0974 · 4.98 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To identify risk factors for failure of anticoagulant thromboprophylaxis in critically ill patients in the ICU. Multivariable regression analysis of thrombosis predictors from a randomized thromboprophylaxis trial. Sixty-seven medical-surgical ICUs in six countries. Three thousand seven hundred forty-six medical-surgical critically ill patients. All patients received anticoagulant thromboprophylaxis with low-molecular-weight heparin or unfractionated heparin at standard doses. Independent predictors for venous thromboembolism, proximal leg deep vein thrombosis, and pulmonary embolism developing during critical illness were assessed. A total of 289 patients (7.7%) developed venous thromboembolism. Predictors of thromboprophylaxis failure as measured by development of venous thromboembolism included a personal or family history of venous thromboembolism (hazard ratio, 1.64; 95% CI, 1.03-2.59; p = 0.04) and body mass index (hazard ratio, 1.18 per 10-point increase; 95% CI, 1.04-1.35; p = 0.01). Increasing body mass index was also a predictor for developing proximal leg deep vein thrombosis (hazard ratio, 1.25; 95% CI, 1.06-1.46; p = 0.007), which occurred in 182 patients (4.9%). Pulmonary embolism occurred in 47 patients (1.3%) and was associated with body mass index (hazard ratio, 1.37; 95% CI, 1.02-1.83; p = 0.035) and vasopressor use (hazard ratio, 1.84; 95% CI, 1.01-3.35; p = 0.046). Low-molecular-weight heparin (in comparison to unfractionated heparin) thromboprophylaxis lowered pulmonary embolism risk (hazard ratio, 0.51; 95% CI, 0.27-0.95; p = 0.034) while statin use in the preceding week lowered the risk of proximal leg deep vein thrombosis (hazard ratio, 0.46; 95% CI, 0.27-0.77; p = 0.004). Failure of standard thromboprophylaxis using low-molecular-weight heparin or unfractionated heparin is more likely in ICU patients with elevated body mass index, those with a personal or family history of venous thromboembolism, and those receiving vasopressors. Alternate management or incremental risk reduction strategies may be needed in such patients.
    Critical Care Medicine 12/2014; 43(2). DOI:10.1097/CCM.0000000000000713 · 6.31 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Repeated therapeutic plasma exchange (TPE) has been advocated to remove heparin-induced thrombocytopenia (HIT) IgG antibodies prior to cardiac/vascular surgery in patients who have serologically-confirmed acute or subacute HIT; for this situation, a negative platelet activation assay (e.g., platelet serotonin-release assay [SRA]) has been recommended as the target serological endpoint to permit safe surgery. We compared reactivities in the SRA and an anti-PF4/heparin IgG-specific enzyme-immunoassay (EIA), testing serial serum samples in a patient with recent (subacute) HIT who underwent serial TPE pre-cardiac surgery, as well as for 15 other serially-diluted HIT sera. We observed that post-TPE/diluted HIT sera-when first testing SRA-negative-continue to test strongly positive by EIA-IgG. This dissociation between the platelet activation assay and a PF4-dependent immunoassay for HIT antibodies indicates that patients with subacute HIT undergoing repeated TPE prior to heparin re-exposure should be tested by serial platelet activation assays even when their EIAs remain strongly positive. Copyright © 2014 American Society of Hematology.
    Blood 11/2014; 125(1). DOI:10.1182/blood-2014-07-590844 · 10.45 Impact Factor
  • John W. Eikelboom · Theodore E. Warkentin
    Journal of the American College of Cardiology 11/2014; 64(20). DOI:10.1016/j.jacc.2014.09.014 · 16.50 Impact Factor
  • Lori-Ann Linkins · Theodore E Warkentin
    Thrombosis Research 11/2014; 135(1). DOI:10.1016/j.thromres.2014.10.019 · 2.45 Impact Factor
  • T E Warkentin · J-A I Sheppard · J C Manheim
    Thrombosis and Haemostasis 10/2014; 112(6). DOI:10.1160/TH14-08-0711 · 4.98 Impact Factor
  • Journal of cardiothoracic and vascular anesthesia 08/2014; 29(5). DOI:10.1053/j.jvca.2014.04.029 · 1.46 Impact Factor
  • Theodore E Warkentin
    [Show abstract] [Hide abstract]
    ABSTRACT: In this issue of Journal of Clinical Monitoring and Computing, Perbet et al. [1] report 2 intensive care unit (ICU) patients with hemofiltration filter clotting associated with heparin-induced thrombocytopenia (HIT), an immune-mediated prothrombotic adverse drug reaction [2, 3]. In both patients, early (within 6 h) hemofilter thrombosis occurred at a time when the patients were not thrombocytopenic. The authors propose that hemofilter clotting should be considered a thrombotic event in the 4Ts scoring system (discussed subsequently), even in the absence of thrombocytopenia. The authors specifically suggest [1] that rapid hemofiltration clotting should score as 2 points for the category, Thrombosis, in this scoring system.The 4Ts [4, 5] is a widely used pretest probability scoring system for diagnosis of HIT [6]. Its value is especially useful in ruling out HIT when the score is low, as this is indicative of a high negative predictive value (i.e., HIT is a most unlikely diagnosis) [7]. ...
    Journal of Clinical Monitoring and Computing 07/2014; 29(1). DOI:10.1007/s10877-014-9594-2 · 1.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Data on the frequency of anti-platelet factor 4/heparin (PF4/H) antibodies and their association with outcomes in intensive care unit (ICU) patients are sparse. In this prospective, observational study we screened 320 consecutive surgical/medical ICU patients for anti-PF4/H antibodies by enzyme-immunoassay (EIA) for immunoglobulin (Ig)G/A/M separately and heparin-induced platelet activation assay (HIPA) at ICU admission (=baseline), day 6, and day 10. HIPA-positive patients were additionally tested by serotonin-release assay (SRA). Patients tested positive by day 10: for anti-PF4/H-IgG = 17.2 % and for anti-PF4/H-IgM = 42.1 %. Within the first 10 ICU days, platelet counts decreased to <100 Gpt/L in 27.8 % patients. However, only seven patients (2.2 %) experienced a drop in the platelet count ≥50 % beginning after the fourth ICU day. These included the only two patients (0.6 %; 95 % confidence interval 0.08-2.2 %) with heparin-induced thrombocytopenia (HIT). Only strong reactions in the HIPA were reproducible by SRA. This study confirms that testing for anti-PF4/H IgG antibodies should be restricted to ICU-patients who develop a platelet count decrease of >50 % that begins after day four of heparin treatment (which may have started before ICU admission). Among patients testing positive by IgG-specific EIA a functional platelet activation assay should be performed (regarding only strong reactions as positive).
    Journal of Thrombosis and Thrombolysis 07/2014; 39(1). DOI:10.1007/s11239-014-1105-2 · 2.17 Impact Factor

Publication Stats

17k Citations
2,340.27 Total Impact Points


  • 1989–2015
    • McMaster University
      • • Department of Medicine
      • • Department of Pathology and Molecular Medicine
      Hamilton, Ontario, Canada
  • 2012
    • University of Pennsylvania
      • Perelman School of Medicine
      Philadelphia, PA, United States
  • 2005–2011
    • University of Greifswald
      • Institute of Immunology and Transfusion Medicine
      Griefswald, Mecklenburg-Vorpommern, Germany
  • 2007
    • Drexel University
      • Division of Pulmonary, Critical Care and Sleep Medicine
      Filadelfia, Pennsylvania, United States
  • 2000–2006
    • Hamilton Health Sciences
      Hamilton, Ontario, Canada
  • 2003
    • The Chinese University of Hong Kong
      • Department of Anaesthesia and Intensive Care
      Hong Kong, Hong Kong
  • 1996
    • St. Joseph's Healthcare Hamilton
      Hamilton, Ontario, Canada