TR-05500; No of Pages 4 Thrombosis Research xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Thrombosis Research journal homepage: www.elsevier.com/locate/thromres

Regular Article

A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients☆ Maheswari Senthil a, Preeti Chaudhary b, David D. Smith c, Patrick E. Ventura c, Paul H. Frankel c, Vinod Pullarkat b, Vijay Trisal a,⁎ a b c

Department of Surgery, City of Hope Medical Center, Duarte, California 91010 Division of Hematology, University of Southern California Keck School of Medicine, Los Angeles, California 90033 Division of Biostatistics, City of Hope Medical Center, Duarte, California 91010

a r t i c l e

i n f o

Article history: Received 12 December 2013 Received in revised form 8 April 2014 Accepted 15 April 2014 Available online xxxx Keywords: Activated partial thromboplastin time Venous thromboembolism Central venous catheters Factor VIII Fibrinogen Cancer

a b s t r a c t Introduction: Hypercoagulability due to high coagulation factor levels resulting from host inflammatory response to cancer contributes to an increased risk of venous thromboembolism (VTE) in cancer patients. Central venous catheters (CVCs) further heighten this risk. Activated partial thromboplastin time (aPTT) can be used to broadly screen for elevated levels of relevant coagulation factors. Our objective was to determine if a shortened aPTT ratio (coagulation time of test- to- reference plasma) was a predictor of CVC-associated VTE in cancer patients. Materials and Methods: We performed a retrospective case–control study on cancer patients undergoing tunneled CVC insertion at our center from 1999 to 2006 and identified 40 patients who had CVC-associated VTE. VTE was confirmed with color duplex ultrasonography or computed tomography scan. For each case, we obtained 5 controls that had the same cancer diagnosis and were matched on the following factors: age, chemotherapy, hormone therapy (if applicable), tobacco use, TNM staging and year of diagnosis. All patients had aPTT testing within 30 days prior to surgery. We compared aPTT and aPTT ratio between cases and controls using Wilcoxon two sample test. Results: aPTT ratio was significantly shorter in patients with CVC-related VTE as compared to controls [0.86 (95% confidence interval (CI) 0.78, 0.94) vs. 0.98 (0.94, 1.01), p = 0.0003]. Mean aPTT was also significantly shorter. [25.6 seconds (95% CI 23.2, 27.9) vs. 28.1 (26.9, 29.3), p = 0.001] aPTT ratios of the controls tended to spread across larger aPTT ratio values whereas those of cases tended to clustered around the mean. Conclusions: Cancer patients undergoing catheter placement who develop CVC-associated VTE have a shorter aPTT and aPTT ratio than those who do not develop VTE. aPTT, a simple and inexpensive test might be useful as a predictor of CVC-associated VTE risk in cancer patients. © 2014 Published by Elsevier Ltd.

Introduction Patients with cancer are at 4- to 6-fold increased risk of developing venous thromembolism (VTE) [1]. This risk varies with cancer type, stage, host responses, chemotherapy, hormonal therapy, surgery and use of central venous catheters (CVCs) [2–5]. The exact mechanism of hypercoagulability in cancer is not completely understood. Host inflammatory reaction to cancer seems to play an important role. Tumor infiltrating macrophages, key mediators of anti-tumor response produce a Abbreviations: VTE, Venous thromboembolism; CVCs, Central venous catheters; aPTT, Activated partial thromboplastin time; CT, Computed tomography; CI, Confidence interval; (F1.2), Prothrombin fragment 1.2; (TAT), Thrombin-antithrombin; FVIII: C, FVIII coagulant activity; LETS, Leiden Thrombophilia Study; CRP, C-reactive protein. ☆ Previous presentation of the manuscript: Presented in part at the Pacific Coast Surgical Association Meeting 2010. ⁎ Corresponding author at: Department of Surgery, City Of Hope, 1500 East Duarte Road, Duarte, CA 91010. Tel.: +1 626 471 7100; fax: +1 626 301 8865. E-mail address: [email protected] (V. Trisal).

number of cytokines which in turn increase the synthesis and release of acute phase proteins like fibrinogen and Factor VIII (FVIII) tilting the hemostatic balance in favor of a prothrombotic state [6]. In addition, elevated levels of coagulation factors particularly FVIII and fibrinogen have recently emerged as independent risk factors for VTE [7–9]. These ‘acute phase reactant’ coagulation factors (FVIII and fibrinogen) can be cumulatively assessed by activated partial thromboplastin time (aPTT), a common laboratory test performed prior to surgical interventions. We reasoned that elevated levels of one or more of the relevant coagulation factors would shorten aPTT and a short aPTT ratio (coagulation time of test-to-reference plasma) can be used as a surrogate marker for elevated levels of these factors in the ‘intrinsic’ coagulation pathway and thereby, predict the risk of postoperative CVC-associated VTE in cancer patients. Our primary predictor of interest was aPTT as aPTT would be shortened by increased levels of both FVIII and fibrinogen, both acute phase reactants and known risk factors for VTE. The spread of normal aPTT values is also sufficiently wide to allow for accurate determination of aPTT shortening. To test our hypothesis we conducted

http://dx.doi.org/10.1016/j.thromres.2014.04.022 0049-3848/© 2014 Published by Elsevier Ltd.

Please cite this article as: Senthil M, et al, A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients, Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.022

2

M. Senthil et al. / Thrombosis Research xxx (2014) xxx–xxx

a retrospective case–control study in cancer patients undergoing tunneled catheter insertion who have had CVC-associated VTE in the post operative period in order to determine if a short aPTT at preoperative screening in this high risk population is a risk factor for the same. Materials and methods Study design Approval for the study was obtained from the Institutional Review Board at City of Hope. We carried out a retrospective case–control study on cancer patients undergoing CVC placement at our center from 1999 to 2006 and developed an episode of post-operative CVCassociated VTE within 30 days after surgery. For the purpose of this study CVC refers to an implantable catheter that have a subcutaneous tunnel such as Hickmans or Port- A- Caths. Peripherally inserted central catheters or percutaneous directly centrally placed catheters (femoral or subclavian lines) were excluded. Patients were identified using available electronic databases with ICD-9 codes for VTE. From these cases, we manually reviewed the charts of eligible patients with CVC-associated VTEs. VTE was diagnosed by color duplex ultrasonography or computed tomography (CT) scan. Cases were defined as patients with aPTT testing performed within 30 days prior to surgery and identified through the medical services system as having had a VTE after catheter placement (excluding VTE 30 days after removal of catheter). Patients who were pre-treated with anticoagulants or those who had major surgeries up to a year prior of the CVC placement were excluded. Among the cases (N = 40), there were 3 patients who had major surgeries prior to CVC placement. One patient had gastro-esophagectomy (4.5 years prior to catheter placement), another patient had nephrectomy with lymph node dissection (3 years prior), and the third patient had sigmoid resection (2 years prior); these are not counting patients with incisional biopsies. The predictors of interest for the study were aPTT and aPTT ratio prior to CVC placement. As a comparator group to the cases, we queried our cancer registry for subjects in the same time frame and similar diagnoses. Among the 8487 eligible controls, we excluded patients who did not have aPTT testing done prior to surgery and those who were pre-treated with anticoagulants.

Table 1 Demographics of the cases of catheter-related venous thromboembolism and matched cases.

Age at Dx Mean ± SE Days between PT labs & catheter insertion Mean ± SE Gender Female Tobacco Use No Yes Former smokers Unknown Site Blood/Bone Marrow Breast Kidney/Prostate/Bladder Skin Colon/Rectum/Anus Esophagus Lymph nodes Ovary Other**

Cases (N = 40)

Controls (N = 157)

43 ± 1.5

48 ± 2.4

8.4 ± 1.3

5.5 ± 0.6*

17

(44)

69

(44)

22 5 9 4

(55) (13) (22) (10)

83 15 32 27

(53) (10) (20) (17)

17 3 2 1 2 1 10 1 3

(43) (7) (5) (3) (5) (3) (25) (3) (7)

72 12 10 3 6 3 36 2 13

(46) (8) (6) (2) (4) (2) (23) (1) (8)

were balanced across clinical factors. However, there was a significant difference between cases and controls for the number of days from when aPTT testing was obtained and when the catheter was inserted (p = 0.021). Coagulation studies

Results

Fig. 1 Panel A demonstrates barplots comparing aPTT between cases with CVC-related VTE and their matched controls. The Wilcoxon test for this means comparison was statistically significant (p = 0.001). We performed an ANOVA analysis to adjust for the difference in days between aPTT testing and catheter insertion. The difference in days was not a significant factor (p = 0.42) however, the difference between cases and controls remained significant (p = 0.03). The mean aPTT was 25.6 seconds (s) (95% confidence interval (CI) 23.2 – 27.9 s) for cases and 28.1 s (95% CI 26.9 – 29.3 s) for controls respectively. Levene test for variance homogeneity was not significant. (p = 0.27). Fig. 1 Panel B demonstrates barplots comparing aPTT ratio between the cases and controls. Consistent with the aPTT results, there were statistically significant differences between the aPTT ratios of cases as compared to controls (p = 0.0003). As with aPTT, an ANCOVA analysis to adjust for the difference in days between coagulation lab testing and catheter insertion was performed. The difference in days was not a significant factor (p = 0.91) however, the difference between cases and controls remained significant (p = 0.0084). The mean aPTT ratio was 0.86 (95% CI 0.78-0.94) for cases and 0.98 (95% CI 0.94-1.01) for controls respectively. From the barplots, there appears to be variance heterogeneity between the aPTT ratios. Specifically, the aPTT ratios of the controls tended to be spread across larger aPTT ratio values whereas the cases tended to be clustered around the mean (Fig. 2). Levene’s test for variance homogeneity was significant (p = 0.039), which may indicate that VTE cases with shorter aPTT ratios are distributed around a mean at or below 1.0. None of the VTE cases were noted to have aPTT ratio of 1.2 or greater. Among the cases, 6/40 (15%) had aPTT ratio of 1.0 or 1.1. In contrast, 17/157 (11%) of the control cases had aPTT ratio of 1.2 or greater, and 70/157 (45%) of the controls had aPTT ratio of 1.0 or greater. Two cases and zero controls had aPTT ratio of 0.6.

Patient characteristics

Discussion

Table 1 shows the baseline characteristics of the 40 cases and 157 matched controls. Due to the matching algorithm, the cases and controls

Our results show that a shortened aPTT ratio prior to CVC insertion in cancer patients is a strong predictor for risk of CVC-associated VTE in the

Statistical methods Our matching algorithm was based on the Mahalanobis distance between cases and controls on the following clinical and demographic factors: age, disease site, chemotherapy, hormone therapy (if applicable), tobacco use, TNM stages and year of diagnosis. In the case of treatment types and disease staging, these factors were matched within each disease site so that the staging and treatments were appropriate for each specific site. For Mahalanobis distance and matching implementation, we used the optmatch package [10]. From the Mahalanobis distances and resulting confusion matrix, each eligible control was scored with each case. Controls were only used once for matching; subjects matched once were not reused. For each case, we obtained 5 controls. Final data set contained 40 cases and 157 matched controls. Our statistical analysis tested the hypothesis that aPTT values and aPTT ratio are shorter in subjects who developed CVC-associated VTE within 30 days of catheter insertion. We compared cases versus controls with respect to these two outcomes using the Wilcoxon two sample test. Barplots were relative frequency charts rounded to the nearest tenth for both aPTT and aPTT ratio.

Please cite this article as: Senthil M, et al, A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients, Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.022

M. Senthil et al. / Thrombosis Research xxx (2014) xxx–xxx

3

Thrombosis cases (N = 40) Matched controls (N = 157) p = 0.0010

Relative Frequency of aPTT Times

0.5

0.4

0.3

0.2

0.1

Probability density function of aPTT Ratio

A

Thrombosis cases (N = 40) Matched controls (N = 157) Levene's homoscedasticity: p = 0.039

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

0.0 15

20

25

30

35

40

45

50

aPTT Time (seconds)

B Thrombosis cases (N = 40) Matched controls (N = 157) p = 0.0003

Relative Frequency of aPTT Ratios

0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0.6 0.7 0.8 0.9

1

1.1 1.2 1.3 1.4 1.5 1.6

aPTT Ratio Fig. 1. Comparison of aPTT and aPTT ratio between cases of post surgical CVC associated VTE and matched controls. (A) Barplots comparing aPTT times between catheter-associated VTE cases and matched controls. The mean aPTT times (95% CI) were 25.6 seconds (s) (23.2 – 27.9 s) and 28.1 s (26.9 – 29.3 s) for cases and controls, respectively. The Wilcoxon test for the aPTT means comparison was statistically significant. (B) Barplot comparing cases and controls with respect to aPTT ratios. The mean aPTT ratios (95% CI) were 0.86 (0.78-0.94) and 0.98 (0.94-1.01) for cases and controls, respectively. The Wilcoxon test for the aPTT ratio means comparison was statistically significant. Variance heterogeneity between the aPTT ratios was determined. aPTT ratios of the controls tended to spread across larger aPTT ratio values whereas the cases tended to clustered around the mean.

immediate post operative period. Although not generally recommended as a screen for bleeding risk after surgery, aPTT as well as ratio are often obtained preoperatively for a variety of reasons. Therefore these results are readily available in most cancer patients undergoing various surgeries and could be considered in making decisions regarding anticoagulation in the post operative period especially if patients have CVCs. aPTT is an inexpensive and simple test that globally explores all coagulation factors except factor VII and seems suitable to reflect any pro-coagulant imbalance resulting from increased levels of coagulation factors, many of which have been associated with a heightened risk for VTE [7–9,11–15]. In a recent study, a shortened aPTT ratio was identified to be an independent risk factor for VTE among patients referred for thrombophilia testing after developing VTE. Results were significant

0.4

0.6

0.8

1.0

1.2

1.4

1.6

aPTT Ratio Fig. 2. Kernal density estimates of aPTT ratio illustrating the spread differences between cases of post surgical CVC associated VTE and matched controls. The Levene test for equal variances was statistically significant. Homoscedasticity refers to the property of having equal statistical variances.

for a 5-fold increased relative risk of VTE in patients with smaller aPTT ratio (≤ 0.8) when compared to patients with greatest aPTT ratios (N1.00) [16]. In another study Gallus et al. performed fourteen tests of hemostasis and fibrinolysis before and after elective surgery and found patients who developed postoperative thrombosis had significantly shorter aPTT than patients without thrombosis. Analysis of the VTE risk at varying levels of aPTT demonstrated that a higher aPTT was protective against developing post-operative thrombosis [17]. The underlying mechanism for increased thrombotic risk with shortened aPTT has not been completely elucidated. What is known is that short aPTT is associated with hemostatic activation, as evidenced by elevations of prothrombin fragment 1.2 (F1.2), thrombin-antithrombin (TAT) complexes, d-dimers and FVIII coagulant activity. (FVIII: C), all markers for increased thrombin generation in plasma [18,19]. Prospective studies have also demonstrated that patients with preoperative evidence of thrombin generation are at significantly higher risk of postsurgical thrombotic complications [20]. The association of shortened aPTT and CVC –associated VTE ratio in the present study is most likely a reflection of elevated FVIII levels in patients with active cancer which may be further elevated post operatively as a result of inflammation. FVIII is an acute phase reactant and plasma levels of FVIII increase in patients with inflammation, cancer or pregnancy. Several studies have supported the high prevalence of FVIII levels in patients with VTE and recognized it to be a dose-dependent phenomenon [9,12,13,15,21]. Furthermore, elevated FVIII levels are a risk factor for recurrent VTE [7,13]. Raised FVIII is most probably the cause of VTE and not simply an epiphenomenon of the underlying cancer associated inflammatory process as no definitive correlation between the plasma levels of FVIII and C-reactive protein (CRP), a sensitive indicator of acute phase reaction has been reported [7,22]. When adjusted for CRP, high factor VIII: C levels still increase the thrombosis risk 6-fold which is very similar to the risk before correction for CRP levels [12]. Our study does not provide data on the exact mechanism of aPTT shortening in cancer patients. It would be useful in future studies to explore the relative contribution of factors that could lead to a shortening of aPTT in cancer patients and their association with post operative VTE. These could include ABO blood group (a determinant of vWF level and ristocetin cofactor activity), ristocetin cofactor activity, FVIII and fibrinogen levels as well as acute phase reactants like sensitive

Please cite this article as: Senthil M, et al, A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients, Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.022

4

M. Senthil et al. / Thrombosis Research xxx (2014) xxx–xxx

CRP measurements. Nonetheless, FVIII has been recognized as an independent risk factor for VTE even after confounding variables are adjusted for [9]. Such studies may provide additional or more accurate predictors of VTE risk, both CVC-related and otherwise in cancer patients undergoing surgical procedures. Conclusion Prophylactic anticoagulation practices for cancer patients with CVCs are not standardized and vary widely. Our results show that there is an association between CVC-associated VTE in the post operative period and shortened aPTT and ratio prior to catheter insertion. This association between shortened aPTT and VTE risk may apply to other forms of VTE as well although this was not examined in the present study. Although there is considerable overlap in both the aPTT and aPTT ratios among cases and controls, there is evidence that aPTT ratios of cancer patients who experienced a CVC-associated VTE tend to be clustered around a mean that is strictly less than 1.0. Analyses such as logistic regression with extensive cross-validation may lead to development of a VTE risk classification model incorporating a shortened aPTT ratio; although any model making such claims should be confirmed in a prospective study. If our results are confirmed in such studies, a shortened aPTT ratio could be a consideration in choosing prophylactic anticoagulation practices for cancer patients undergoing CVC insertion. Authors disclosures of potential conflicts of interest The authors have no conflicts of interest to declare References [1] Thodiyil PA, Kakkar AK. Variation in relative risk of venous thromboembolism in different cancers. Thromb Hemost 2002;87(6):1076–7. [2] Baron JA, Gridley G, Weiderpass E, Nyrén O, Linet M. Venous thromboembolism and cancer. Lancet 1998;351(9109):1077–80. [3] Callander N, Rapaport SI. Trousseau’s syndrome. West J Med 1993;158:364–71. [4] Bona RD. Thrombotic complications of central venous catheters in cancer patients. Semin Thromb Hemost 1999;25:145–55. [5] Rickles FR, Levine MN. Epidemiology of thrombosis in cancer. Acta Haematol 2001;106:6–12.

[6] Sutherland DE, Weitz IC, Liebman HA. Thromboembolic complications of cancer: epidemiology, pathogenesis, diagnosis, and treatment. Am J Hematol 2003;72(1):43–52. [7] Kyrle PA, Minar E, Hirschl M, Bialonczyk C, Stain M, Schneider B, et al. High plasma levels of factor VIII and the risk of recurrent venous thromboembolism. N Engl J Med 2000;343(7):457–62. [8] Koster T, Rosendaal FR, Reitsma PH, van der Velden PA, Briët E, Vandenbroucke JP. Factor VII and fibrinogen levels as risk factors for venous thrombosis: a casecontrol study of plasma levels and DNA polymorphism. The Leiden Thrombophilia Study (LETS). Thromb Haemost 1994;71(6):719–22. [9] Koster T, Blann AD, Briët E, Vandenbroucke JP, Rosendaal FR. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet 1995;345(8943):152–5. [10] Hansen BB, Fredrickson M. optmatch: Functions for optimal matching, version 0.8-1. [http://www.r-project.org, accessed May 2010], In: R Core Team, editor. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2013 [http://www.R-project.org]. [11] Meijers JC, Tekelenburg WL, Bouma BN, Bertina RM, Rosendaal FR. High levels of Factor XI as a risk factor for venous thrombosis. N Engl J Med 2000;342(10):696–701. [12] Kamphuisen PW, Eikenboom JC, Vos HL, Pablo R, Sturk A, Bertina RM, et al. Increased levels of factor VIII and fibrinogen in patients with venous thrombosis are not caused by acute phase reactions. Thromb Haemost 1999;81(5):680–3. [13] Kraaijenhagen RA, in't Anker PS, Koopman MM, Reitsma PH, Prins MH, van den Ende A, et al. High plasma concentration of factor VIII:C is a major risk factor for venous thromboembolism. Thromb Haemost 2000;83(1):5–9. [14] Van Hylckama A, van der Linden IK, Bertina RM, Rosendaal FR. High levels of factor IX increases the risk of venous thrombosis. Blood 2000;95(12):3678–82. [15] Ota S, Yamada N, Ogihara Y, Tsuji A, Ishikura K, Nakamura M, et al. High plasma level of factor VIII: an important risk factor for venous thromboembolism. Circ J 2011;75(6):1472–5. [16] Tripodi A, Chantarangkul V, Martinelli I, Bucciarelli P, Mannucci PM. A shortened activated partial thromboplastin time is associated with the risk of venous thromboembolism. Blood 2004;104(12):3631–4. [17] Gallus AS, Hirsh J, Gent M. Relevance of preoperative and postoperative blood tests to postoperative leg-vein thrombosis. Lancet 1973;2(7833):805–9. [18] Ten Boekel E, Bartels P. Abnormally short activated partial thromboplastin times are related to elevated plasma levels of TAT, F1 + 2, D-dimer and FVIII:C. Pathophysiol Haemost Thromb 2002;32(3):137–42. [19] Korte W, Clarke S, Lefkowitz JB. Short activated partial thromboplastin times are related to increased thrombin generation and an increased risk for thromboembolism. Am J Clin Pathol 2000;113(1):123–7. [20] Falanga A, Ofosu FA, Cortelazzo S, Delaini F, Consonni R, Caccia R, et al. Preliminary study to identify cancer patients at high risk of venous thrombosis following major surgery. Br J Haematol 1993;85(4):745–50. [21] O'Donnell J, Tuddenham EG, Manning R, Kemball-Cook G, Johnson D, Laffan M. High prevalence of elevated factor VIII levels in patients referred for thrombophilia screening: role of increased synthesis and relationship to the acute phase reaction. Thromb Haemost 1997;77(5):825–8. [22] Tichelaar V, Mulder A, Kluin-Nelemans H, Meijer K. The acute phase reaction explains only a part of initially elevated factor VIII:C levels: a prospective cohort study in patients with venous thrombosis. Thromb Res 2012;129(2):183–6.

Please cite this article as: Senthil M, et al, A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients, Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.022

A shortened activated partial thromboplastin time predicts the risk of catheter-associated venous thrombosis in cancer patients.

Hypercoagulability due to high coagulation factor levels resulting from host inflammatory response to cancer contributes to an increased risk of venou...
277KB Sizes 0 Downloads 0 Views