Prediction of Venous Thromboembolism in Patients

VOLUME

29

䡠

NUMBER

15

䡠

MAY

20

2011

JOURNAL OF CLINICAL ONCOLOGY

O R I G I N A L

R E P O R T

Prediction of Venous Thromboembolism in Patients With Cancer by Measuring Thrombin Generation: Results From the Vienna Cancer and Thrombosis Study Cihan Ay, Daniela Dunkler, Ralph Simanek, Johannes Thaler, Silvia Koder, Christine Marosi, Christoph Zielinski, and Ingrid Pabinger From the Medical University of Vienna, Austria. Submitted September 21, 2010; accepted January 19, 2011; published online ahead of print at www.jco.org on April 4, 2011. Supported by a grant from the Jubila¨umsfonds of the Austrian National Bank (project numbers 10935 and 12739) and by an unrestricted grant from Pfizer Austria. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Ingrid Pabinger, MD, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria; e-mail: [email protected]. © 2011 by American Society of Clinical Oncology 0732-183X/11/2915-2099/$20.00 DOI: 10.1200/JCO.2010.32.8294

A

B

S

T

R

A

C

T

Purpose Patients with cancer are at increased risk of venous thromboembolism (VTE). Laboratory tests measuring the overall thrombophilic tendency might be useful to assess VTE risk. The aim of this study was to investigate thrombin generation, a key process in hemostasis, as predictor of cancer-associated VTE. Patients and Methods The Vienna Cancer and Thrombosis Study (CATS) is a prospective observational cohort study of patients with newly diagnosed cancer or progression of disease after remission. The study end point is occurrence of objectively confirmed symptomatic or fatal VTE within a follow-up period of 2 years. Thrombin generation was measured with a commercially available assay. Results One thousand thirty-three patients with malignancies of the breast (n ⫽ 151), lung (n ⫽ 148), upper (n ⫽ 44) and lower gastrointestinal tract (n ⫽ 125), pancreas (n ⫽ 67), kidney (n ⫽ 34), prostate (n ⫽ 122), and brain (n ⫽ 134) or lymphoma (n ⫽ 126), multiple myeloma (n ⫽ 26), and other tumor types (n ⫽ 56) were observed for a median observation period of 517 days. VTE occurred in 77 patients (7.5%). Patients with elevated peak thrombin (defined as values ⱖ 611 nM thrombin, representing the 75th percentile of the total study population) had an increased risk of VTE with a hazard ratio of 2.1 (95% CI, 1.3 to 3.3, P ⫽ .002) in multivariable analysis including age, sex, surgery, chemotherapy, and radiotherapy. The cumulative probability of developing VTE after 6 months was significantly higher in patients with elevated peak thrombin than in those with lower peak thrombin (11% v 4%; log-rank test: P ⫽ .002). Conclusion Measurement of thrombin generation may help identify patients with cancer at high risk of VTE. J Clin Oncol 29:2099-2103. © 2011 by American Society of Clinical Oncology

INTRODUCTION

The presence of cancer represents an acquired prothrombotic condition, and thus patients with cancer are at high risk of venous thromboembolism (VTE).1-3 However, the pathogenesis of VTE in cancer is complex and principally depends on a number of interactions of cancer cells with the clotting system, which lead to a hypercoagulable state. Cancer cells themselves activate blood coagulation directly or indirectly in several ways. For instance, they are able to release procoagulants, to activate endothelial cells, leukocytes, and platelets by cytokines and to produce a factor X-activating cystein protease, mucinous glycoproteins, and circulating tissue factor– bearing microparticles.2,4,5 These biologic factors lead to the generation of thrombin and

subsequent fibrin formation. Interestingly, thrombin together with tissue factor (TF) is also believed to be a key mediator in establishing the pathopysiological link between cancer and thrombosis.1 Furthermore, patient- and treatment-related factors, such as immobilization, comorbidities, or anticancer treatments, also contribute to the prothrombotic state and the risk of cancer-associated VTE.6 To overcome the complex pathogenesis and the multiple risk factors for cancer-associated VTE, single laboratory tests reflecting a multifactorial thrombophilic state could help assess the overall risk of VTE in patients with cancer. As the generation of thrombin plays a central role in the coagulation system, the measurement of the thrombin generation potential provides a promising method for quantifying the composite effect of the multiple factors © 2011 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at Bibliothek der MedUniWien (51061) on September 17, Copyright © 2011 American of Clinical Oncology. All rights reserved. 2011Society from 149.148.252.245

2099

Ay et al

determining the coagulation capacity and the influence of the environment on these risk factors.7-9 Functional assays have been developed that allow an efficient measurement of the thrombin generation potential.10 These in vitro tests use TF and phospholipids to activate the coagulation cascade. The concentration of generated thrombin is registered over time, resulting in a thrombin generation curve, from which various parameters can be determined that describe thrombin activity, such as the lag phase (time until thrombin burst), peak amount of thrombin generation, time to peak thrombin, velocity of thrombin generation, or the total amount of thrombin generated (area under the thrombin generation curve [AUC]). Recent clinical research has focused on the measurement of thrombin generation and its predictive value for VTE. Thrombin generation has been demonstrated to be generally increased in subjects at risk of VTE. In patients without cancer, an increased thrombin generation potential has been found to be associated with first and recurrent events of VTE.11-16 The clinical relevance of thrombin generation for predicting cancer-associated VTE is yet unknown. Considering the hypercoagulable state in patients with cancer and the high incidence of VTE in this population, we hypothesized that measurement of thrombin generation could help stratify patients with cancer into high- and low-risk groups for occurrence of VTE. To test this hypothesis, we measured thrombin generation with a fluorogenic assay in 1,033 patients with cancer enrolled in a prospective observational cohort study and investigated the association between thrombin generation and the occurrence of VTE. PATIENTS AND METHODS Patients and Study Design The study population comprises patients recruited into the Vienna Cancer and Thrombosis Study (CATS), an ongoing prospective observational cohort study initiated in October 2003 at the Medical University of Vienna with approval of the local ethics committee and in accordance with the Declaration of Helsinki. The detailed methodology of CATS and exact inclusion and exclusion criteria for this study have been reported in detail previously.17-19 Briefly, the study includes cancer patients with newly diagnosed cancer or disease progression after complete or partial remission who have not recently received chemotherapy (within the past 3 months), radiotherapy, or surgery (within the past 2 weeks). All patients give their written informed consent before enrollment. The following cancer sites, which have to be confirmed by histology, are eligible for CATS: brain, breast, lung, stomach, colorectal, pancreas, kidney, prostate, or other sites (mainly of the gynecological system and sarcoma), and hematologic malignancies (myeloma and lymphoma). Patients are observed prospectively over a 2-year observation period, until occurrence of VTE or death, loss of follow-up, or withdrawal of consent. At the time of study inclusion venous blood samples were drawn for laboratory analyses. The study population consists of 1,033 patients enrolled between October 2003 and December 2008, in whom all clinical and laboratory parameters and follow-up information were available. Outcome Measure The main outcome measure of the study was occurrence of VTE, either symptomatic or fatal VTE, confirmed by duplex sonography, phlebography, and/or computed tomography or autopsy within 2 years after study inclusion, as described previously.17-19 Laboratory Analysis Venous blood was collected in plasma vacuum tubes (Vacuette, GreinerBio One, Kremsmuenster, Austria; containing 1/10 volume sodium citrate stock solution at 0.129 mmol/L) by sterile and atraumatic venipuncture. Samples were centrifuged to obtain platelet-poor plasma and aliquots were stored 2100

© 2011 by American Society of Clinical Oncology

at ⫺80 degrees until the time of analysis. Thrombin generation was measured with a commercially available assay kit (Technothrombin TGA kit, Techonoclone, Vienna, Austria) on a fully automated, computer-controlled microplate reader (BioTek, FL ⫻800) and a specially adapted software (Technothrombin TGA, Vienna, Austria) using the fluorogenic substrate Z-Gly-Gly-Arg-AMC (Bachem, Bubendorf, Switzerland) according to the manufacturer’s instructions and as previously described.20 The reaction was triggered with the TGA RC low reagent, which contained 71.6 pM recombinant human tissue factor lipidated in 3.2 ␮mol/L phospholipid micelles (phosphatidylcholine [2.56 ␮mol/L] and phosphatidylserine [0.64 ␮mol/L]). For analyzing the association of thrombingenerationwithVTEinpatientswithcancer,theparameterpeakthrombin (the maximum concentration of thrombin generation) was used. Statistical Analysis Continuous variables were described by the median and the 25th and 75th percentile; nominal variables were described by absolute numbers and percentages. The median follow-up time was calculated with the reverse Kaplan-Meier method. The correlations between D-dimer, prothrombin fragment 1 ⫹ 2, coagulation factor VIII, soluble P-selectin, hemoglobin, platelet and leukocyte count, and peak thrombin were described by the Spearman correlation coefficient. Of main interest was the association between peak thrombin and the risk of developing VTE. Peak thrombin was analyzed as a continuous variable and as a binary variable with a threshold at 611 nM thrombin. This cutoff point, representing the 75th percentile of peak thrombin of the total study population, was predefined for this study. The area under the receiver operating characteristic (ROC) curve given by c-statistic was 0.611. Univariate and multivariable Cox regression analyses was used for calculating the risk of VTE from study inclusion until first VTE, last follow-up, patient’s death, or maximal length of follow-up of 2 years. Multivariable Cox regression analysis comprised the following parameters: peak thrombin, age, sex, surgery, chemotherapy, and radiotherapy. The covariate of main interest was peak thrombin. Hazard ratios (HR) of the continuous peak thrombin are given per 100 nM increase. We assumed that surgery, chemotherapy, and radiotherapy would entail a modified risk for VTE not only at the exact time point of the procedure, but even for a certain period afterward. Therefore, three timedependent dichotomous variables were included in the statistical model that indicated times of possible influence on the VTE risk by surgery (from the day of surgery plus 6 consecutive weeks), chemotherapy (from the first day of a treatment cycle until the last day plus 4 weeks), or radiotherapy (from the first until the last day of a treatment plus 4 weeks). Furthermore, the analysis was adjusted for age at study inclusion and sex. In a second multivariable Cox regression analysis, we additionally adjusted for tumor site and stage by introducing four groups of patients into the statistical model ((1) brain tumors, (2) hematologic malignancies, (3) solid tumors without distant metastasis, and (4) solid tumors with distant metastasis). The Cox regression models were tested for all pairwise interactions and interactions with log(time) by means of candidate variables. As no significant interaction was found (P value smaller than .01), no interaction was added to the final model. Kaplan-Meier analysis was used to visualize the risk of developing VTE in patients with peak thrombin below and above the threshold. A log-rank test was used to assess differences in developing VTE in these two groups. A P value smaller than .05 was regarded as statistically significant. All calculations were conducted with SAS 9.2.

RESULTS

Characteristics of Patients Baseline characteristics of the investigated cohort of 1,033 patients are presented in Table 1. In total, 77 VTE events (7.5%) were recorded during a median observation period of 517 days (25th to 75th percentile: 237 to 731). The cumulative probability of VTE was 5.8% after 6 months and 7.9% after 1 year. The characteristics of patients with cancer with VTE, classification of cancer and site of cancer are given in Table 2. JOURNAL OF CLINICAL ONCOLOGY


Thrombin Generation and Risk of Cancer-Associated VTE

Table 1. Baseline Demographic Characteristics of the Total Study Population (n ⫽ 1,033), Site of Cancer, and Treatment During Observation No.

% 62 53-68

572 461

55.4 44.6

134 151 148 44 125 67 34 122 26 126 56

13.0 14.7 14.4 4.3 12.1 6.5 3.3 11.8 2.5 12.2 5.4

693 397 480

67.1 38.4 56.5

Thrombin Generation and Risk of VTE Parameters of thrombin generation in the study population were as follows (median [25th to 75th percentile]): lag phase 9 [7-11] minutes, peak thrombin 500 [360-611] nM thrombin, time to peak 12.5 [10.5-15.5] minutes, velocity index of thrombin generation 129 [73-185] nM/min, AUC 4,386 [3,816-4,890] nM thrombin). Patients who developed VTE had statistically significantly higher peak thrombin values than those without VTE events during follow-up (556 [432-677] v 499 [360-603] nM, P ⫽ .014), they also had a shorter lag phase (8 [7-10] v 9 [7-11] minutes, P ⫽ .036), a shorter time to peak thrombin (11.5 [9.5-14] v 13 [11-16] minutes, P ⫽ .006) and a higher velocity index (160 [111-223] v 127 [71-183] nM/min, P ⫽ .004). No statistically significant difference was observed for the AUC (4,475 [4,087-4,915] v 4,386 [3,804-4.890] nM thrombin, P ⫽ .197) in patients who developed VTE compared to those without VTE during observation. Peak thrombin was the variable of interest for further statistical analyses. The cumulative probability of developing VTE after 6 months was significantly higher in patients with elevated peak TG (defined as values ⱖ 611 nM thrombin, representing the 75th percentile of the total study population) than in those with lower peak thrombin (11% v 4%, P ⫽ .002). Figure 1 compares the Kaplan-Meier estimates of the VTE risk among patients with elevated peak thrombin (ⱖ 611 nM thrombin) and nonelevated peak thrombin (⬍ 611 nM thrombin). In univariate Cox regression analysis peak thrombin showed a statistically significant association with VTE per 100 nM increase (HR, 1.15; 95% CI, 1.02 to 1.30; P ⫽ .020); and this association was still seen in multivariable analysis after adjustment for age, sex, surgery, chemotherapy, and radiotherapy (HR, 1.15; 95% CI, 1.02 to 1.29; P ⫽ .019). When peak thrombin was dichotomized into groups below and above the 75th percentile of the total study population (cutoff point: www.jco.org

Characteristic

No.

Median age at study entry, years 25th-75th percentile Sex Female Male Site of thrombotic event Isolated DVT of the lower extremity Isolated PE Fatal PE Combined DVT of the lower extremity and PE Inferior caval vein thrombosis Portal vein thrombosis Combined DVT of the lower extremity and portal vein thrombosis Sinus vein thrombosis Brachial vein thrombosis Combined PE and brachial vein thrombosis Jugular vein thrombosis Site of cancer Brain Breast Lung Stomach Colorectal Pancreas Kidney Prostate Multiple myeloma Lymphoma Other

% 62 53-68

31 46

40.3 59.7

34 27 3 2 1 3

44.2 35.1 3.9 2.6 1.3 3.9

1 1 2 1 2

1.3 1.3 2.6 1.3 2.6

20 3 7 7 11 12 1 2 1 9 4

26.0 3.9 9.1 9.1 14.3 15.6 1.3 2.6 1.3 11.7 5.2

Abbreviations: VTE, venous thromboembolism; DVT, deep vein thrombosis; PE, pulmonary embolism.

611 nM thrombin), the HR of patients with high peak thrombin (peak thrombin ⱖ 611 nM) was significantly increased in univariate analysis (HR, 2.0; 95% CI, 1.3 to 3.2; P ⫽ .003). This association remained

0.25

Probability of VTE

Characteristic Median age at study entry, years 25th-75th percentile Sex Female Male Site of cancer Brain Breast Lung Stomach Colorectal Pancreas Kidney Prostate Multiple myeloma Lymphoma Other Cancer treatment during observation period Chemotherapy Surgery Radiotherapy

Table 2. Demographic and Clinical Characteristic of Patients With Cancer Who Developed VTE During the Observation Period (n ⫽ 77)

Peak TG ≥ 75th percentile Peak TG < 75th percentile

0.20 0.15

0.10

0.05

0

100

No. of patients at risk 217 691

200

300

400

500

600

700

115 370

107 330

94 272

Time (days) 180 589

157 484

136 409

Fig 1. Cumulative probability of venous thromboembolism (VTE) in all patients (n ⫽ 1,033). Patients with increased peak thrombin (peak TG; cutoff: 611 nM thrombin, blue line) are compared to those with nonincreased peak TG (gold line).



2101

Ay et al

similarly increased in multivariable analysis including age, sex, surgery, chemotherapy, and radiotherapy (HR, 2.1; 95% CI, 1.3 to 3.3; P ⫽ .002). In a second multivariable analysis the association of high peak thrombin with the occurrence of VTE was additionally adjusted for different tumor subgroups (considering tumor site and tumor stage) and still remained statistically significant (HR, 1.9; 95% CI, 1.2 to 3.0; P ⫽ .008). No relevant correlations were observed between peak thrombin and D-dimer (r ⫽ 0.20), prothrombin fragment 1⫹2 (r ⫽ 0.16), coagulation factor VIII (r ⫽ 0.20), soluble P-selectin (r ⫽ 0.11), hemoglobin (r ⫽ ⫺0.16), platelet (r ⫽ 0.26), and leukocyte count (r ⫽ 0.17). DISCUSSION

VTE in patients with cancer is a frequent and serious complication that renders patient management more difficult and leads to increased morbidity and mortality. Although VTE is a preventable disease and prophylactic anticoagulation is available, it still remains a major challenge to stratify patients with cancer into high- or low-risk groups for VTE. In our study, we have found that an increased thrombin generation potential, which was shown by high levels of peak thrombin is independently associated with an increased risk of patients with cancer to develop VTE. This association was present in the analyses of peak thrombin as a continuous as well as a dichotomized variable. In multivariable analysis the risk of VTE was two-fold higher in patients with a high peak thrombin (peak thrombin levels ⱖ 611 nM) than in those with lower levels. The association of peak thrombin with the VTE risk proved to be significant in multivariable regression analysis after adjustment for treatment regimen, age, and sex. The probability of developing VTE after 6 months was 11% in patients with cancer with high peak thrombin levels compared to 4% in those with lower levels. Accordingtocurrentinternationalguidelinesprimarythromboprophylaxis in patients with cancer is recommended only in certain high-risk settings such as during hospitalization or after major cancer surgery.21,22 In the ambulatory setting, primary thromboprophylaxis is not recommended, with the exception of patients with multiple myeloma treated with thalidomide or lenalidomide. Only recently, Agnelli et al23 presented a randomized placebo-controlled trial of a low-molecular-weight heparin that showed a benefit of primary thromboprophylaxis for preventing VTE in ambulatory patients receiving chemotherapy for metastatic or locally advanced cancer. However, a clear benefit of thromboprophylaxis was restricted to only a subgroup of patients with cancer and the authors of this study suggested targeting thromboprohylaxis at those patients considered to be at highest risk of VTE by means of risk stratification. In line with this study, in two other trials, both of which are currently available only as meeting abstracts, routine primary thromboprophylaxis with lowmolecular-weightheparinswasreportedtoreducesymptomaticVTE rates in patients with advanced pancreatic cancer, a cancer type known for its very high VTE risk.24,25 One way to individualize thromboprophyaxis is the identification of patients at high risk of VTE by deploying of a previously published risk scoring model using clinical and laboratory parameters.26 This risk scoring system developed by Khorana et al, which incorporated the site of cancer, platelet count, hemoglobin, and/or use of erythropoiesis-stimulating agents, leukocyte count, and BMI, predicted for symptomatic VTE during chemotherapy. Only recently our group 2102


validated the predictive value of the Khorana risk model in a broader range of patients with cancer and proposed an expanded risk model, which additionally includes two biomarkers (soluble P-selectin and D-Dimer) and provides greater accuracy in predicting the risk of VTE.27 Another possibility to stratify patients according to their VTE risk would be through measurement of biomarkers or laboratory tests that globally reflect a prothrombotic or hypercoagulable state. In this respect, in vitro thrombin generation measurement is a new and appealing method that allows determination of an individual’s coagulation potential. Thrombin generation has been widely investigated in clinical studies for its predictive value for first and recurrent episodes of VTE in patients without cancer. These studies indicate that an increased thrombin generation potential may predict occurrence and recurrence of VTE. Our study is the first that specifically tests the value of thrombin generation for predicting VTE in patients with cancer and identifies high peak thrombin levels as a novel risk factor for cancer-associated VTE. Thrombin generation assays evaluate an overall hemostatic status of patients at risk of VTE and measure the cumulative effect of prothrombotic tendencies.7-10 Therefore, they may be particularly useful for prediction of VTE in patients with cancer, as the VTE risk in patients with cancer is multifactorial and increases with the number of risk factors. The measurement of thrombin generation alone or together with other clinical and laboratory parameters incorporated in risk scoring models may also contribute to a betterstratificationofpatientswithcancerintohigh-orlow-riskgroupsof VTE. Interestingly, in vivo parameters that indirectly indicate thrombin generation, such as the prothrombin fragment 1⫹2, thrombin antithrombin complex, fibrinopeptid A, and D-dimer have previously been reported to be increased in patients with cancer compared to patients without cancer and support an activation of blood coagulation in patients with cancer.28,29 Prospective studies, including previous reports from CATS, have shown that activation of blood coagulation, as reflected by in vivo parameters of thrombin generation (eg, prothrombin fragment 1⫹2 and D-dimer), may be predictive of the occurrence of VTE in patients with cancer18 and also recurrence of cancer-associated VTE and poor survival.30,31 We also calculated correlation coefficients of peak thrombin, prothrombin fragment 1⫹2, and D-dimer and found no relevant correlation between these parameters. Therefore, we suggest that peak thrombin may add additional information to thrombotic risk assessments in patients with cancer. Several assays have been developed to measure the in vitro thrombin generation in plasma. Using the information from the thrombin generation curve that is produced by these assays, thrombin generation can be expressed in different ways.10,11 The most appropriate parameter, which best describes the thrombin activity in the assay that we applied in our study, is peak thrombin (ie, the maximal thrombin concentration). Another frequently used parameter is the endogenous thrombin potential that determines the AUC by calibrated automated thrombography. Although both methods are commonly reported in the literature,7,10 comparison between studies is difficult, since variations in preanalytic variables, tissue factor, and phospholipid concentrations can cause significant changes. However, the assays have been found to be reproducible within a single laboratory when the same reagent and method for measurement of thrombin generation is used.7,10 Generally, further standardizations of the thrombin generation assays are required to minimize interlaboratory variations, in order to conduct future studies for defining normal values of thrombin generation in the general population and for JOURNAL OF CLINICAL ONCOLOGY


Thrombin Generation and Risk of Cancer-Associated VTE

assessing the clinical utility of thrombin generation assays when identifying a hyper- or hypocoagulable phenotype. The main limitation of our study is that only one blood sample for measurement of thrombin generation was drawn at study inclusion. Due to organizational and logistical reasons, it was not possible to perform serial measurements during the follow-up period. Furthermore, as autopsies were not performed in all cases, the number of fatal events might be underestimated. However, an independent adjudication committee examined all cases, confirmed the diagnosis, and proved the clinical significance of all events. In conclusion, our prospective study provides data on a large number of patients with cancer with different tumor entities over a relatively long time period. Using a relatively simple assay that globally reflects an individual’s coagulation potential, we were able to identify patients with cancer at increased risk of VTE. Whether assessment of thrombin generation may be clinically useful for an individual stratification of patients with cancer according to their VTE risk needs to be elucidated in further large prospective and interventional studies. REFERENCES 1. Rickles FR, Patierno S, Fernandez PM: Tissue factor, thrombin, and cancer. Chest 124:58S-68S, 2003 2. Falanga A, Panova-Noeva M, Russo L: Procoagulant mechanisms in tumour cells. Best Pract Res Clin Haematol 22:49-60, 2009 3. Wun T, White RH: Venous thromboembolism (VTE) in patients with cancer: Epidemiology and risk factors. Cancer Invest 27:63-74, 2009 (suppl 1) 4. Lyman GH, Khorana AA: Cancer, clots and consensus: New understanding of an old problem. J Clin Oncol 27:4821-4826, 2009 5. Bick RL: Cancer-associated thrombosis. N Engl J Med 349:109-111, 2003 6. Lip GY, Chin BS, Blann AD: Cancer and the prothrombotic state. Lancet Oncol 3:27-34, 2002 7. Baglin T: The measurement and application of thrombin generation. Br J Haematol 130:653-661, 2005 8. Pabinger I, Ay C: Biomarkers and venous thromboembolism. Arterioscler Thromb Vasc Biol 29:332-336, 2009 9. Hemker HC, Al Dieri R, De Smedt E, et al: Thrombin generation, a function test of the haemostaticthrombotic system. Thromb Haemost 96:553-561, 2006 10. van Veen JJ, Gatt A, Makris M: Thrombin generation testing in routine clinical practice: Are we there yet? Br J Haematol 142:889-903, 2008 11. Lutsey PL, Folsom AR, Heckbert SR, et al: Peak thrombin generation and subsequent venous thromboembolism: The Longitudinal Investigation of Thromboembolism Etiology (LITE) study. J Thromb Haemost 7:1639-1648, 2009 12. van Hylckama Vlieg A, Christiansen SC, Luddington R, et al: Elevated endogenous thrombin potential is associated with an increased risk of a first deep venous thrombosis but not with the risk of recurrence. Br J Haematol 138:769-774, 2007 13. Hron G, Kollars M, Binder BR, et al: Identification of patients at low risk for recurrent venous

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest.

AUTHOR CONTRIBUTIONS Conception and design: Cihan Ay, Christoph Zielinski, Ingrid Pabinger Financial support: Cihan Ay, Ingrid Pabinger Administrative support: Cihan Ay, Silvia Koder Provision of study materials or patients: Christine Marosi, Christoph Zielinski Collection and assembly of data: Cihan Ay, Ralph Simanek, Johannes Thaler, Christine Marosi Data analysis and interpretation: Cihan Ay, Daniela Dunkler, Silvia Koder, Ingrid Pabinger Manuscript writing: All authors Final approval of manuscript: All authors

thromboembolism by measuring thrombin generation. JAMA 296:397-402, 2006 14. Tripodi A, Legnani C, Chantarangkul V, et al: High thrombin generation measured in the presence of thrombomodulin is associated with an increased risk of recurrent venous thromboembolism. J Thromb Haemost 6:1327-1333, 2008 15. Eichinger S, Hron G, Kollars M, et al: Prediction of recurrent venous thromboembolism by endogenous thrombin potential and D-dimer. Clin Chem 54:2042-2048, 2008 16. Besser M, Baglin C, Luddington R, et al: High rate of unprovoked recurrent venous thrombosis is associated with high thrombin-generating potential in a prospective cohort study. J Thromb Haemost 6:1720-1725, 2008 17. Ay C, Simanek R, Vormittag R, et al: High plasma levels of soluble P-selectin are predictive of venous thromboembolism in cancer patients: Results from the Vienna Cancer and Thrombosis Study (CATS). Blood 112:2703-2708, 2008 18. Ay C, Vormittag R, Dunkler D, et al: D-dimer and prothrombin fragment 1 ⫹ 2 predict venous thromboembolism in patients with cancer: Results from the Vienna Cancer and Thrombosis Study. J Clin Oncol 27:4124-4129, 2009 19. Vormittag R, Simanek R, Ay C, et al: High factor VIII levels independently predict venous thromboembolism in cancer patients: The cancer and thrombosis study. Arterioscler Thromb Vasc Biol 29:2176-2181, 2009 20. Ay L, Kopp HP, Brix JM, et al: Thrombin generation in morbid obesity: Significant reduction after weight loss. J Thromb Haemost 8:759-765, 2010 21. Lyman GH, Khorana AA, Falanga A, et al: American Society of Clinical Oncology guideline: Recommendations for venous thromboembolism prophylaxis and treatment in patients with cancer. J Clin Oncol 25:5490-5505, 2007 22. Geerts WH, Bergqvist D, Pineo GF, et al: Prevention of venous thromboembolism: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (ed 8). Chest 133:381S-453S, 2008

23. Agnelli G, Gussoni G, Bianchini C, et al: Nadroparin for the prevention of thromboembolic events in ambulatory patients with metastatic or locally advanced solid cancer receiving chemotherapy: A randomised, placebo-controlled, double-blind study. Lancet Oncol 10:943-949, 2009 24. Riess H, Pelzer U, Deutschinoff G, et al: A prospective, randomized trial of chemotherapy with or without the low molecular weight heparin (LMWH) enoxaparin in patients (pts) with advanced pancreatic cancer (APC): Results of the CONKO 004 trial. J Clin Oncol 27:203s, 2009 (suppl; abstr LBA4506) 25. Maraveyas A, Waters J, Roy R, et al: Gemcitabine with or without prophylactic weight-adjusted dalteparin (WAD) in patients with advanced or metastatic pancreatic cancer (APC): A multicentre, randomised phase IIB trial (the UK FRAGEM study). Thromb Res 125:S161-5 (OC-02), 2010 (suppl 2) 26. Khorana AA, Kuderer NM, Culakova E, et al: Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 111:4902-4907, 2008 27. Ay C, Dunkler D, Marosi C, et al: Prediction of venous thromboembolism in cancer patients. Blood 116:5377-5382, 2010 28. Goldenberg N, Kahn SR, Solymoss S: Markers of coagulation and angiogenesis in cancer-associated venous thromboembolism. J Clin Oncol 21:41944199, 2003 29. Edwards RL, Rickles FR, Moritz TE, et al: Abnormalities of blood coagulation tests in patients with cancer. Am J Clin Pathol 88:596-602, 1987 30. Sallah S, Husain A, Sigounas V, et al: Plasma coagulation markers in patients with solid tumors and venous thromboembolic disease receiving oral anticoagulation therapy. Clin Cancer Res 10:72387243, 2004 31. Cosmi B, Legnani C, Cini M, et al: The role of D-dimer and residual venous obstruction in recurrence of venous thromboembolism after anticoagulation withdrawal in cancer patients. Haematologica 90:713-715, 2005

■ ■ ■

www.jco.org



2103