Elevated Fibrin-Related Markers in Patients with Malignant Diseases Frequently Associated with Disseminated Intravascular Coagulation and Venous Thromboembolism

□ ORIGINAL ARTICLE □

Elevated Fibrin-related Markers in Patients with Malignant Diseases Frequently Associated with Disseminated Intravascular Coagulation and Venous Thromboembolism

Yoshiki Yamashita 1,HideoWada2, Hideki Nomura 1, Toshiro Mizuno 1, Kanako Saito 1, Norikazu Yamada 3, Kunihiro Asanuma 4, Masanobu Usui 5, Yuki Kamimoto 6, Takeshi Matsumoto 7, Kohshi Ohishi 7 and Naoyuki Katayama 1

Abstract

Objective Many patients with malignant diseases are frequently complicated with some type of thrombosis, such as venous thromboembolism (VTE) or disseminated intravascular coagulation (DIC). Methods This retrospective study was designed to examine the frequency of thrombosis in 478 patients with malignant diseases in comparison to that observed in 121 patients without malignant diseases and to evaluate the efficacy of fibrin-related markers (FRMs), such as soluble fibrin, fibrinogen and fibrin degradation products and D-dimer, in diagnosing thrombosis. Results The frequency of thrombosis, including 62 cases of VTE, 63 cases of DIC and nine cases of cere- brovascular thrombosis, was significantly higher in the patients with malignant diseases (28.0%) than in the patients without malignant diseases (12.5%). DIC was frequently detected in the patients with hepatic cell cancer and hematopoietic malignancy, while VTE was frequently observed in the patients with colon cancer, breast cancer and urinary tract cancer. The FRMs levels were significantly higher in the patients with thrombosis than in the patients without thrombosis. A receiver operating characteristic analysis showed these markers to be useful for diagnosing thrombosis. Conclusion Patients with malignant diseases have a high risk of thrombosis, and elevated FRMs levels are useful for diagnosing thrombosis in patients with malignant diseases.

Key words: malignant disease, DIC, VTE, D-dimer, SF

(Intern Med 53: 413-419, 2014) (DOI: 10.2169/internalmedicine.53.1102)

creased by concomitant patient-related thrombotic risk fac- Introduction tors, such as an advanced age, infection, heart disease, respi- ratory disease, hospitalization and surgical or nonsurgical Patients with cancer or hematopoietic malignancy have a cancer treatment (3, 4). The association between thrombotic high risk of developing thrombotic diseases, such as venous conditions and malignant disease is known as Trousseau’s thromboembolism (VTE) (1) or disseminated intravascular syndrome, which has multiple definitions and mecha- coagulation (DIC) (2). These risks vary according to the nisms (5). The hyperexpression or release of tissue factor type of malignancy and stage of disease, and are steadily in- (TF) is one of the most important factors for the develop-

１Departments of Hematology and Oncology, Mie University Graduate School of Medicine, Japan, ２Departments of Molecular and Laboratory Medicine, Mie University Graduate School of Medicine, Japan, ３Department of Cardiology, Mie University Graduate School of Medicine, Japan, ４Department of Orthopedic Surgery, Mie University Graduate School of Medicine, Japan, ５Departments of Hepatobiliary Pancreatic and Trans- plant Surgery, Mie University Graduate School of Medicine, Japan, ６Departments of Obstetrics and Gynecology, Mie University Graduate School of Medicine, Japan and ７Department of Blood Transfusion, Mie University Graduate School of Medicine, Japan Received for publication June 2, 2013; Accepted for publication September 29, 2013 Correspondence to Dr. Hideo Wada, [email protected]

413 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102

Table 1. Subjects

Malignant diseases Benign diseases p value Age 65.0 (55.0-72.3) 49.0 (37.0-63.0) p<0.001 Sex (F:M) 200 : 278 67 : 54 NS Thrombosis 134 14 p<0.001 Cerebral vascular event 9 0 NS Disseminated intravascular coagulation 63 4 p<0.01 Venous thromboembolism 62 10 NS

ment of hypercoagulability due to thrombosis in patients bosis due to symptoms or abnormalities on laboratory tests with malignant disease (6, 7). (Table 1). The malignant diseases consisted of 48 cases of The presence of thrombosis is sometimes fatal and can hematopoietic malignancy, 52 cases of hepatic cell carci- disturb the quality of life among patients with cancer and noma, 99 cases of lung cancer, 26 cases of stomach cancer, leukemia. Therefore, providing thromboprophylaxis is im- 21 cases of uterine cancer, 98 cases of colon cancer, 24 portant for these patients (8). As there are many patients cases of ovarian cancer, 13 cases of breast cancer, 16 cases with cancer and leukemia, all patients cannot be treated with of urinary tract cancer,18 cases of cancer of unknown ori- anticoagulants. Therefore, it is necessary to assess the pres- gin, 18 cases of sarcoma, 14 cases of pancreatic cancer and ence of hypercoagulability in patients with malignant dis- 31 others. The cases without malignant diseases consisted of eases, and only patients with hypercoagulability and those at 42 cases of soft tissue tumors (STTs), 18 cases of bone tu- high risk for thrombosis should be treated with anticoagu- mors, 13 cases of thymoma, eight cases of fibroid tumors, lants. The plasma levels of fibrin-related markers (FRMs), seven cases of lipoma, seven cases of essential thrombo- such as D-dimer, fibrinogen and fibrin degradation products cythemia (ET), five cases of neck tumors, five cases of (FDPs) and soluble fibrin (SF), are useful for diagnosing schwannoma, three cases of angioma, two cases of ovarian thrombosis and have been reported to be elevated in patients tumors, two cases of meningioma and nine others. ET some- with deep vein thrombosis (DVT)/pulmonary embolism times increases the levels of marked blast cells, resulting in [PE; (9, 10) and DIC (10, 11). An unlikely pretest probabil- a poor outcome; however, the seven ET patients had few ity in combination with a negative D-dimer test has been re- blast cells in the peripheral blood, suggesting that they were ported to be able to safely rule out DVT in Europe and in the chronic phase of the disease. None of the patients North America (12). The cut-off level for D-dimer as a with ET had VTE and were treated with warfarin. negative predictor of DVT is reported to be less than 0.5 μg/ The plasma concentrations of D-dimer and soluble fibrin mL (13). The plasma levels of D-dimer and SF may in- (SF) and the serum levels of fibrinogen and fibrin degrada- crease in cancer patients without thrombosis. Increasing the tion products (FDP) were retrospectively examined in all pa- cut-off value of D-dimer in cancer patients has thus been re- tients and analysed in order to identify any correlations with ported to possibly increase the test’s clinical usefulness (14), the diagnosis of thrombosis. The study protocol was ap- while the combination of a low or unlikely Wells pretest proved by the Human Ethics Review Committee of Mie probability with a negative D-dimer result can be used to University Hospital, and informed consent was obtained rule out DVT in patients with cancer (15). from all subjects. A total of 148 patients were diagnosed to The present study was designed to examine the incidence have thrombosis, while 451 were not. No patients were ex- of thrombotic diseases in patients with malignant diseases amined after undergoing liver transplantation or within three and to evaluate the cut-off values of FRMs for thrombotic days after an operation. DVT was diagnosed using echo or diseases. For this purpose, the FRM levels were assessed in venography. PE was diagnosed using either ventilation- 478 patients suspected of having thrombosis in comparison perfusion lung scanning, computed tomography (CT) or pul- to those observed in 121 patients without malignant dis- monary angiography. DIC was diagnosed according to the eases. International Society on Thrombosis and Haemostasis (ISTH) overt-DIC diagnostic criteria (16). Cerebral thrombo- Materials and Methods sis was diagnosed on CT or magnetic resonance imaging (MRI). The concentrations of SF, FDP and D-dimer were meas- Subjects ured in the patients with thrombosis at the onset of disease Between August 1, 2003 and July 31, 2010, 478 consecu- and in those without thrombosis at the first consultation. tive patients with advanced malignant diseases (median age: The data were usually obtained before the administration of 65.0 years, 25-75% range: 55.0-72.3 years old, and sex: 200 chemotherapy or anticoagulation therapy. In each case, the women and 278 men) and 121 consecutive patients without consultation was conducted within two days after the ap- malignant diseases (49.0 years; 37.0-63.0 years, 67 women pearance of either symptoms or laboratory abnormalities. and 54 men) were suspected of having some type of throm-

414 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102

Table 2. Frequency of Thrombosis

n DIC DVT Thrombosis Hematopoietic tumor 48 10 (20.8%) 4 (8.3%) 16 (33.3%) Hepatic cell carcinoma 52 22 (42.3%) 4 (7.7%) 29 (55.8%) Lung cancer 99 5 (5.1%) 9 (9.1%) 15 (9.1%) Stomach cancer 26 5 (19.2%) 2 (7.7%) 8 (30.8%) Uterus cancer 21 0 (0%) 9 (42.9%) 9 (42.9%) Colon cancer 98 2 (2.0%) 10 (19.4%) 13 (13.3%) Ovarian cancer 24 2 (8.3%) 9 (37.5%) 11 (45.8%) Breast cancer 13 3 (23.1%) 2 (15.4%) 5 (38.5%) Urinary tract cancer 16 4 (25.0%) 3 (18.8%) 8 (50.0%) Unknown origin cancer 18 1 (5.6%) 1 (5.6%) 2 (11.1%) Sarcoma 18 3 (16.7%) 0 (0 %) 3 (16.7%) Pancreas cancer 14 2 (14.3%) 3 (21.4%) 5 (35.7%) Others 31 4 (12.9%) 6 (19.49%) 10 (32.3%) Total 478 63(13.2%) 62 (13.0%) 134 (28.0%)

Figure 1. Plasma levels of SF in the patients with and without malignant diseases (MD), MD patients with and without Measurement of the plasma concentrations of D- thrombosis, MD patients with disseminated intravascular co- dimer and SF and the serum levels of FDP agulation (DIC) and MD patients with venous thromboembo- The plasma and serum samples were obtained and ana- lism (VTE). SF: soluble fibrin, MD: malignant diseases, TH: lyzed within four hours. The plasma D-dimer levels were thrombosis, DIC: disseminated intravascular coagulation, measured with LPIA-ACE D-dimer (Mitsubishi Chemical VTE: venous thromboembolism, ***: p< 0.001 Medience Corporation, Tokyo, Japan) using JIF-23 monoclonal antibodies. JIF-23 monoclonal antibodies, which rec- ognize the plasmin-digested N-terminus of the γ chain on cated with DIC, and no patients with DIC were complicated the D region, were used for latex aggregation (17). The SF with symptomatic VTE. In particular, the rate of complica- was also determined according to the latex agglutination tion with DIC was significantly higher in the patients with method using IATRO SF (Mitsubishi Chemical Medience malignant diseases (63; 13.2%) than in the patients without Corporation, Tokyo, Japan) containing the monoclonal anti- malignant diseases (4; 3.6%, p<0.01) (Table 1). DIC was body IF-43, which recognizes a segment of the fibrin Aα frequently detected in the patients with hepatic cell carci- chain [(Aα-17-78) residue segment] exposed in the E region noma, hematopoietic malignancy, stomach cancer, breast of the fibrin monomer (FM) when the FM molecule binds cancer and urinary tract cancer, while VTE was frequently the D region of another FM or fibrinogen. The antibody was observed in the patients with colon cancer, breast cancer, coated for the SF assay (10). The serum FDP concentrations urinary tract cancer, uterine cancer, pancreatic cancer and were measured according to the latex agglutination method ovarian cancer (Table 2). using IATRO FDP (Mitsubishi Chemical Medience Corpora- The levels of SF (4.9 μg/mL; 2.6-8.2 μg/mL vs. 2.9 μg/ tion, Tokyo, Japan). mL; 0.8-5.1 μg/mL, p<0.001), FDP (11.6 μg/mL; 6.0-23.5 μg/mL vs. 4.1 μg/mL; 1.0-10.2 μg/mL, p<0.05), and D- Statistical analysis dimer (4.7 μg/mL; 0.6-10.6 μg/mL vs. 0.6 μg/mL; 0.3-1.8 The data are expressed as the median (25% tile-75% tile). μg/mL, p<0.001) were significantly higher in the patients Differences between groups were examined for significance with malignant diseases than in those with benign diseases using the Mann-Whitney U test. A p-value of less than 0.05 (Fig. 1-3). Among the patients with malignant diseases, the was considered to indicate a significant difference. The use- levels of SF (12.5 μg/mL; 7.1-29.9 μg/mL vs. 4.1 μg/mL; fulness of the D-dimer and SF levels in diagnosing throm- 1.9-6.2 μg/mL, p<0.001), FDP (22.0 μg/mL; 10.5-52.8 μg/ bosis and DVT or PE was examined using a receiver operat- mL vs. 6.0 μg/mL; 1.7-11.2 μg/mL, p<0.001) and D-dimer ing characteristic (ROC) analysis (18). The cut-off values (14.1 μg/mL; 8.1-23.2 μg/mL vs. 1.1 μg/mL; 0.5-6.6 μg/mL, were determined according to a ROC analysis. All statistical p<0.001) were significantly higher in the patients with analyses were performed using the SPSS II software pack- thrombosis than in those without thrombosis (Fig. 1-3). age (SPSS Japan, Tokyo, Japan). These markers were also significantly higher in the patients with DIC or VTE than in those without thrombosis, al- Results though only the D-dimer levels were significantly higher in the patients with DIC than in those with VTE (17.0 μg/mL; The frequency of thrombosis, including 62 cases of VTE, 9.0-30.2 μg/mL vs. 12.0 μg/mL; 7.1-17.4 μg/mL, p<0.05). 63 cases of DIC and nine cases of cerebral thrombosis, was The levels of SF were significantly higher in the patients significantly higher in the patients with malignant diseases with hepatic cell cancer, the levels of FDP were signifi- (134; 28.0%) than in the patients without malignant disease cantly higher in the patients with hepatic cell cancer, hema- (14; 12.5%, p<0.001). No patients with VTE were compli- topoietic malignancy, ovarian cancer, breast cancer, urinary

415 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102

Figure 2. Plasma levels of FDP in the patients with and with- Figure 3. Plasma levels of D-dimer in the patients with and out malignant diseases (MD), MD patients with and without without malignant diseases (MD), MD patients with and with- thrombosis, MD patients with disseminated intravascular co- out thrombosis, MD patients with disseminated intravascular agulation (DIC) and MD patients with venous thromboembo- coagulation (DIC) and MD patients with venous thromboem- lism (VTE). FDP: fibrinogen and fibrin degradation products, bolism (VTE). MD: malignant diseases, TH: thrombosis, DIC: MD: malignant diseases, TH: thrombosis, DIC: disseminated disseminated intravascular coagulation, VTE: venous throm- intravascular coagulation, VTE: venous thromboembolism, boembolism, ***: p<0.001, *: p<0.05 ***: p< 0.001, *: p< 0.05

Table 3. Plasma Levels of SF, FDP and D-dimer in the Patients with Various Malignant Diseases

SF (ȝg/mL) FDP (ȝg/mL) D-dimer (ȝg/mL) Hematopoietic tumor 8.6 (0.9-20.5) 12.4 (6.4-22.9)** 7.3 (3.2-11.0)*** Hepatic cell carcinoma 10.4 (6.0-23.6)*** 17.2 (9.5-24.8)*** 14.9 (7.3-22.6)*** Lung cancer 4.8 (3.2-6.3) 9.6 (0.1-13.2) 0.9 (0.4-5.9) Stomach cancer 3.7 (0.7-12.8) 8.2 (0.4-15.1) 1.1 (0.5-8.1) Uterus cancer 5.6 (3.0-6.2) 7.3 (1.8-15.0) 6.5 (4.4-13.1)*** Colon cancer 4.0 (1.5-8.0) 9.8 (3.1-22.2) 0.8 (0.4-3.5) Ovarian cancer 5.1 (2.6-10.1) 13.7 (7.7-26.2)* 7.5 (3.0-11.4)*** Breast cancer 5.6 (4.2-10.9) 17.7 (6.2-140.8)* 7.8 (4.3-19.3)* Urinary tract cancer 5.4 (3.8-7.6) 27.8 (14.3-58.4)*** 7.2 (3.4-27.3)*** Unknown origin cancer 6.1 (3.2-8.9) 11.0 (5.3-20.4) 4.5 (0.5-15.5) Sarcoma 2.9 (1.7-5.6) 32.0 (14.3-58.2)* 3.8 (0.5-15.5) Pancreas cancer 5.8 (3.6-9.3) 22.7 (7.5-36.7)* 10.4 (4.8-13.8)*** Others 2.7 (1.2-6.2) 8.6 (2.4-14.9) 8.2 (1.0-12.7)** Malignant diseases without thrombosis 4.1 (1.9-6.2) 6.0 (1.7-11.2) 1.1 (0.5-6.6) *; p<0.05, **; p<0.01, ***; p<0.001, in comparison to the FDP, SF and D-dimer levels in the non-thrombotic patients with malignant diseases tract cancer, sarcoma or pancreatic cancer and the levels of rates were significantly higher in the patients with thrombo- D-dimer were significantly higher in the patients with he- sis than in those without thrombosis (Table 5). The mortality patic cell cancer, hematopoietic malignancy, uterine cancer, rate was markedly high in the patients with DIC and tended ovarian cancer, breast cancer, urinary tract cancer, sarcoma, to be high in the patients with VTE or cerebral thrombosis. pancreatic cancer or others in comparison to those observed The plasma levels of SF and FDP were significantly higher in the non-thrombotic patients with malignant diseases (Ta- in the 90-day non-survivors than in the 90-day survivors; ble 3). In the ROC analysis, the areas under the curve however, the levels of D-dimer were not significantly differ- (AUCs) for SF, FDP and D-dimer with respect to thrombo- ent (Table 6). sis, DIC and VTE were high (Table 4). The adequate cut-off values of SF, FDP and D-dimer for thrombosis were 6.3 μg/ Discussion mL, 10.6 μg/mL and 7.4 μg/mL, respectively, which were slightly different from those observed for VTE and DIC. This study demonstrated a high frequency of thrombosis The odds ratios for SF and D-dimer were highest in the pa- in patients with malignant diseases in comparison to those tients with DIC, while the odds ratio for VTE was higher in without malignant diseases, suggesting that a hypercoagula- the patients with an elevated D-dimer level than in those ble state exists in patients with malignant diseases. An ele- with an elevated SF value. The 30- and 90-day mortality vated TF expression has been reported in tumor cells and

416 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102

Table 4. ROC Analysis of the SF, FDP and D-dimer Levels Table 5. Outcome of Patients with Thrombosis for Thrombosis, DIC and VTE 30 days mortality 90 days mortality SF FDP D-dimer Patients with thrombosis 26/134 (19.4%)*** 37/134 (27.6%)### AUC thrombosis 0.819 0.855 0.884 DIC 20/63 (31.7%) 29/63 (46.0%) DIC 0.888 0.884 0.909 VTE 5/62 (8.1%) 7/62 (11.3%) VTE 78.5 0.792 0.861 Cerebral thrombosis 1/9 (11.1%) 1/9 (11.1%) Cutoff value thrombosis 6.3 ȝg/mL 10.6 ȝg/mL 7.4 ȝg/mL Patients without thrombosis 12/344 (3.5%)*** 27/344 (8.5%)### DIC 8.1 ȝg/mL 12.1 ȝg/mL 8.0 ȝg/mL ***; p<0.001 in difference between patients with and without thrombosis VTE 6.1 ȝg/mL 9.9 ȝg/mL 6.8 ȝg/mL ###; p<0.001 in difference between patients with and without thrombosis Sensitivity thrombosis 77.1% 73.5% 79.5% DIC 83.6% 75.8% 81.5% VTE 72.1% 70.6% 75.9% Specificity thrombosis 77.1% 73.5% 79.5% DIC 83.6% 75.8% 81.5% VTE 72.1% 70.6% 75.9% Odds ratio thrombosis 10.9 7.6 14.9 DIC 25.5 9.6 19.6 VTE 6.3 6.0 9.9

Table 6. Plasma Levels of SF, FDP and D-dimer in 90 Days Survivor and Non-survivor

SF (ȝg/mL) FDP (ȝg/mL) D-dimer (ȝg/mL) 90 days survivor 2.92 (0.53-8.87)*** 4.60 (2.45-7.41)*** 10.15 (4.50-10.15) 90 days non-survivor 13.00 (7.13-24.34)*** 14.10 (5.12-39.56)*** 16.00 (9.80-28.60) ***; p<0.001 leukocytes obtained from patients with malignant dis- with colon cancer, breast cancer, urinary tract cancer, uterine eases (2, 7). Elevated TF activates coagulation factor VII cancer, pancreatic cancer and ovarian cancer. Colon cancer, thereby causing thrombosis. The difference in the frequency uterine cancer, pancreatic cancer and ovarian cancer are not of DIC was significant between the patients with malignant frequently associated with DIC, suggesting that the onset of diseases and those without malignant diseases; however, that most cases of VTE in patients with these types of malignant of VTE was not. The patients with malignant diseases were diseases depends on the condition of the host, including fac- significantly older than those without malignant diseases, in- tors associated with surgery, anticancer agents, infections, dicating that the age of patients with malignant diseases etc. However, it is difficult to draw any valid conclusions may affect the onset of thrombosis (19). However, the due to the small number of subjects in the different sub- pathogenic mechanisms of DIC depend on the underlying groups, such as those with pancreatic cancer, breast cancer disease, such as leukemia, solid cancer and sepsis (2, 7). and ovarian cancer. These findings suggest that DIC is primarily caused by a The levels of FRMs were significantly higher in the pa- high expression of TF and plasminogen activator (20, 21), tients with malignant diseases than in those without malig- while VTE is usually caused by hemostatic and several nant diseases, thus suggesting that both hypercoagulable and physical factors. Enhancement of both the coagulation and hyperfibrinolytic states exist in patients with malignant dis- fibrinolysis systems may therefore play an important role in eases (6, 7). Elevated D-dimer levels have been reported in the onset of DIC in patients with malignant diseases (22). elderly individuals (23), and many patients with malignant Although no patients had both VTE and DIC, VTE and DIC diseases are elderly, suggesting that the plasma D-dimer lev- have similar pathogenic mechanisms, such as an increased els are slightly high in patients with malignant diseases. In expression of TF. addition, the plasma levels of D-dimer have been reported to In this study, the frequency of DIC was similar to that of be increased in patients with ascites (24). Furthermore, in VTE. VTE and DIC are the most important thrombotic com- the present study, among the patients with malignant dis- plications in patients with malignant diseases, although their eases, the levels of FRMs were significantly higher in the frequency depends on the stage and type of malignancy. patients with thrombosis than in those without thrombosis, DIC was frequently detected in the patients with hepatic cell although only the D-dimer levels were significantly higher carcinoma, hematopoietic malignancy, stomach cancer, in the patients with DIC than in those with VTE. These breast cancer and urinary tract cancer, indicating similar findings indicate that secondary fibrinolysis is more signifi- findings to those of a questionnaire survey conducted by the cant in patients with DIC than in those with VTE and that Japanese Ministry Health and Welfare (7). These types of the D-dimer level is a sensitive marker of secondary fibri- malignancies may be associated with either a high TF level nolysis. There have been many reports regarding the SF and or organ failure, such as that due to liver dysfunction or in- D-dimer levels in patients with DIC and VTE (9, 10). In fection (22). VTE was frequently observed in the patients this study, the levels of SF were significantly high only in

417 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102 the patients with Hepatocellular carcinoma (HCC), suggest- References ing that an elevated SF level is caused by thrombosis rather than malignancy. Meanwhile, the levels of FDP and D-dimer 1. Thaler J, Ay C, Pabinger I. Venous thromboembolism in cancer were significantly high in the patients with various types of patients-risk scores and recent randomised controlled trials. malignant diseases, suggesting that elevated FDP and D- Thromb Haemost 108: 1042-1048, 2012. 2. Wada H, Matsumoto T, Hatada T. Diagnostic criteria and labora- dimer levels are caused by both thrombotic states and malig- tory tests for disseminated intravascular coagulation. Expert Rev nant diseases. Elevated levels of FRMs, such as D-dimer Hematol 5: 643-652, 2012. and SF, are also observed in patients under a hypercoagula- 3. Falanga A. The incidence and risk of venous thromboembolism ble state (25), such as those with DVT (26, 27), DIC (28) or associated with cancer and nonsurgical cancer treatment. Cancer hyperlipidemia (29). Therefore, it is important to determine Invest 27: 105-115, 2009. 4. Nomura H, Wada H, Mizuno T, et al. Elevated fibrin-related mark- the adequate cut-off values of FRMs for diagnosing throm- ers in patients with malignant diseases suspected of having thrombosis in patients with malignant diseases. botic disorders. Clin Appl Thromb Hemost 16: 266-272, 2010. In the ROC analysis, the AUCs of the FRMs for the diag- 5. Varki A. Trousseau’s syndrome: multiple definitions and multiple nosis of thrombosis were significantly high, suggesting that mechanisms. Blood 110: 1723-1729, 2007. FRMs are useful for diagnosing thrombosis in patients with 6. Falanga A, Barbui T, Rickles FR. Hypercoagulability and tissue factor gene upregulation in hematologic malignancies. Semin malignant or other diseases. The adequate cut-off values of Thromb Hemost 34: 204-210, 2008. SF, FDP and D-dimer for thrombosis were 6.3 μg/mL, 10.6 7. Wada H. Disseminated intravascular coagulation. CCA 344: 13- μg/mL and 7.4 μg/mL, respectively, although they were 21, 2004. slightly different among the patients with thrombosis, DIC 8. Akl EA, Terrenato I, Barba M, Sperati F, Muti P, Schünemann HJ. and VTE, indicating that FRMs are not useful for making Extended perioperative thromboprophylaxis in patients with can- the differential diagnosis between VTE and DIC. The cut-off cer. A systematic review. Thromb Haemost 100: 1176-1180, 2008. 9. Linkins LA, Bates SM, Ginsberg JS, Kearon C. Use of different value for SF is similar to that reported in previous stud- D-dimer levels to exclude venous thromboembolism depending on ies (10); however, that for D-dimer was higher in the pa- clinical pretest probability. J Thromb Haemost 2: 1256-1260, tients with malignant diseases than in those with other un- 2004. derlying diseases (approximately 4.0 μg/mL) (10). The odds 10. Wada H, Kobayashi T, Abe Y, et al. Elevated levels of soluble fi- ratios of SF and D-dimer were highest in the patients with brin or D-dimer indicate high risk of thrombosis. J Thromb Hae- most 6: 1253-1258, 2006. DIC, while the odds ratio for VTE was higher in the pa- 11. Wada H, Sakuragawa N, Shiku H. Hemostatic molecular markers tients with an elevated D-dimer level than in those with an before onset of disseminated intravascular coagulation in leukemic increased SF value; however, these odds ratios were not suf- patients. Semin Thromb Hemost 24: 293-297, 1998. ficiently high. 12. Wells PS, Anderson DR, Rodger M, et al. Evaluation of D-dimer Regarding the limitations of this study, the number of pa- in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 349: 1227-1235, 2003. tients was small, and there were various underlying and 13. Houbouyan-Reveillard LL, Mihoubi A, Houdijk WP, et al. Pre- thrombotic diseases. Complications of thrombosis, particu- liminary evaluation of two new rapid immunoturbidimetric D- larly DIC, increase mortality. These findings are similar for dimer assays in patients with clinically suspected venous throm- sepsis (2, 7) and highlight the importance of diagnosing boembolism (VTE). Thromb Haemost 84: 770-774, 2000. thrombosis in patients with malignant diseases. In this study, 14. Righini M, Le Gal G, De Lucia S, et al. Clinical usefulness of D- dimer testing in cancer patients with suspected pulmonary embo- the SF and FDP levels were useful for predicting poor out- lism. Thromb Haemost 95: 715-719, 2006. comes, whereas the D-dimer level was not, suggesting that 15. Carrier M, Lee AY, Bates SM, Anderson DR, Wells PS. Accuracy evaluating the FDP level is useful for diagnosing DIC (2, 7), and usefulness of a clinical prediction rule and D-dimer testing in while evaluating the D-dimer level is useful for diagnosing excluding deep vein thrombosis in cancer patients. Thromb Res VTE (10). 123: 177-183, 2008. 16. Taylor Jr FB, Toh CH, Hoots WK, Wada H, Levi M. Towards In addition, this study was retrospective. The present definition, clinical and laboratory criteria, and a scoring system for study confirmed that measuring the FRMs levels is useful disseminated intravascular coagulation. Thromb Haemost 86: for making the diagnosis of thrombosis in patients with ma- 1327-1330, 2001. lignant diseases; however, the cut-off value for D-dimer 17. Matsuda M, Terukina S, Yamazumi K, Maekawa H, Soe G. A used to diagnose thrombosis is high in comparison to that monoclonal antibody that recognizes the NH2-terminal conforma- tion of fragment D. In: Fibrinogen 4, Current Basic and Clinical observed in patients with other underlying diseases. Aspects. Matsuda M, Iwanaga S, Takada A, Henschen A, Eds. 1990: 43-48. The authors state that they have no Conflict of Interest (COI). 18. Goldstein BJ, Mushlin AI. Use of a single thyroxine test to evaluate ambulatory medical patients for suspected hypothyroidism. J Acknowledgement GenInternMed2: 20-24, 1987. 19. Spencer FA, Ginsberg JS, Chong A, Alter DA. The relationship This study was supported in part by research grants from the between unprovoked venous thromboembolism, age, and acute Japanese Ministry of Health, Labour and Welfare and the Japa- myocardial infarction. J Thromb Haemost 6: 1507-1513, 2008. nese Ministry of Education, Science, Sports and Culture. 20. Wada H, Nagano T, Tomeoku M, et al. Coagulant and fibrinolytic activities in the leukemic cell lysates. Thromb Res 30: 315-322, 1982.

418 Intern Med 53: 413-419, 2014 DOI: 10.2169/internalmedicine.53.1102

21. Wada H, Kumeda Y, Ogasawara Z, et al. Plasminogen activators 1999. and their inhibitors in leukemic cell homogenates. Am J Hematol 26. Minamikawa K, Wada H, Wakita Y, et al. Increased activated pro- 42: 166-170, 1993. tein C-protein C inhibitor complex levels in patients with pulmo- 22. Kaido T, Oe H, Yoshikawa A, Mori A, Arii S, Imamura M. Tissue nary embolism. Thromb Haemost 71: 192-194, 1994. factor is a useful prognostic factor of recurrence in hepatocellular 27. Yamada N, Wada H, Nakase T, et al. Hemostatic abnormalities in carcinoma in 5-year survivors. Hepatogastroenterology 52: 1383- patients with pulmonary embolism compared with that in deep 1387, 2005. vein thrombosis. Blood Coag Fibr 6: 627-633, 1995. 23. Wannamethee SG, Lowe GD, Whincup PH, Rumley A, Walker M, 28. Wada H, Mori Y, Shimura M, et al. Poor outcome in disseminated Lennon L. Physical activity and hemostatic and inflammatory intravascular coagulation or thrombotic thrombocytopenic purpura variables in elderly men. Circulation 105: 1785-1790, 2002. patients with severe vascular endothelial cell injuries. Am J Hema- 24. Saray A, Mesihovic R, Gornjakovic S, et al. Association between tol 58: 189-194, 1998. high D-dimer plasma levels and ascites in patients with liver cir- 29. Wada H, Mori Y, Kaneko T, et al. Elevated plasma levels of vas- rhosis. Med Arh 66: 372-374, 2012. cular endothelial cell markers in patients with hypercholes- 25. DempfleCE.Theuseofsolublefibrin in evaluating the acute and terolemia. Am J Hematol 44: 112-116, 1993. chronic hypercoagulable state. Thromb Haemost 82: 673-683,

419