Hemostatic Derangements in Patients with Solid Malignant Tumors

Patients with solid malignant tumor show an increased susceptibility to thromboembolic events as compared to the general population. Although clinical symptoms occur less frequently, disorders of coagulation are very common in such patients [1]. Thromboembolism is one of the most common causes of death in cancer patients [1]. Ovarian, pancreatic, prostatic and lung cancer, and mucin-producing carcinomas of gastrointestinal tract are among the malignancies most frequently associated with thromboembolic episodes [1]. The overall incidence of clinical thromboembolic disease (TED) in patients with cancer has been reported to vary between 1% and 11% [2]. The incidence of TED in postmortem studies of cancer patients is considerably higher. In one prospective study, Ambrus and associates reported that thrombosis and/or bleeding were the second most common cause of death in hospitalized cancer patients [3]. An abnormal result on routine tests of blood coagulation in cancer patients has been reported to occur in as many as 92% of the patients [4, 5]. A wide range of coagulation disturbances can occur in patients with malignancies, which can predispose patients to hemorrhage or thrombosis [6]. In 1974, the concept of “intravascular coagulation and fibrinolysis syndrome (ICF)” was introduced by Owen and Bowie and the purpose was to examine the incidence and type of hemostatic derangement in patients with cancer [7]. Cooper and associates classified this syndrome into three groups; decompensated which is defined by depressed platelets or fibrinogen levels along with other hemostatic abnormalities, compensated in which platelets or fibrinogen values are normal but other hemostatic tests are abnormally prolonged, and overcompensated whereas the platelets or fibrinogen are increased together with other hemostatic abnormalities. This has also been correlated with the conventional terms; acute, sub-acute and chronic DIC with decompensated, overcompensated and compensated DIC respectively [7, 8]. The aims of the study were to determine the presence of hemostatic derangements in patients suffering from solid malignant tumors. A secondary aim was to elucidate the association of intravascular coagulation and fibrinolysis (ICF) with grade of tumor, histologic type of tumor and presence of metastasis.

This case-controlled study was conducted during a period of six months, from January to July 2004. A total of 40 patients with solid malignant tumors of various organs were included in this study; 13 of these patients were admitted to Al-Yermok Teaching Hospital in Baghdad and 27 patients were from the Hospital of Radiation and Nuclear Medicine in Baghdad. Fifteen age and sex matched healthy volunteers were included in this present study as control group, 9 males and 6 females, the age range was 40-70 years for males and 40-60 years for females. After taking consent, data was collected on a questionnaire including the complete history of the disease, clinical findings on physical examination, with special attention to any evidence of increased bleeding or clotting tendency at enrollment or in the past since the diagnosis of malignancy; details regarding the tumour including grade, histologic type and distant metastasis; and laboratory data. All laboratory tests were performed in Al-Yermok Teaching Hospital Laboratory. We excluded patients with co-morbid conditions that could affect coagulation profile, or patients using certain medication or those with history of coagulation disorders. The platelet count was conducted on blood whereas plasma was isolated for remaining laboratory tests including prothrombin time (PT), activated partial thromboplastin time (aPTT), thrombin time (TT), fibrinogen and D-dimer. Patients were considered to have evidence of intravascular coagulation and fibrinolysis (ICF) syndrome if their plasma D-dimer level was more than 0.5µg/ml. The concept of ICF of Owen and Bowie was adopted and their classification of overcompensated, compensated and decompensated ICF was tested [8, 9] and an attempt was made to identify those key tests that might enable us to discover patients with coagulation problems. The D-dimer and platelets were used as indicators and separated the patients into four groups.

1. Patients with no ICF (normal D-dimer).
2. Those with overcompensated ICF (elevated D-dimer and elevated platelets count).
3. Those with compensated ICF (elevated D-dimer but normal platelets count).
4. Those with decompensated ICF (elevated D-dimer and decreased platelets count).

A battery of hemostatic tests were applied among these four groups, and several tests appeared to be helpful in differentiating those with no ICF, compensated ICF, overcompensated ICF and decompensated ICF. These were PT, APTT, fibrinogen level and TT. In this study, the FDP test which was used by Owen and Bowie study [7] has been replaced by D-dimer test for the diagnosis of ICF syndrome because the D-dimer test is regarded now to be more specific for fibrin degradation products whereas the formation of fibrinolytic degradation products (FDP), X, Y, D and E fragments may be either fibrinogen or fibrin derived following the plasmin digestion [9, 23]. Moreover, the criteria of Owen and Bowie for the diagnosis of ICF was utilized, although in recent years many new criteria were added to diagnose ICF [11, 12, 13] but none of these were applicable in the present study because of the non-availability of some of relevant laboratory tests. Pooled plasma from at least 4 healthy individuals, were prepared and divided into aliquots each contains 1 ml, stored at – 20Cº to be used simultaneously with patients plasma for a period not exceeding 10 days. Positive and negative control plasma (for D-dimer test) was supplied with plasma D-dimer kit. The reference range of the hematological tests were platelet count 150000 – 400000, PT was 12 – 15 seconds, PTT was 32– 39 seconds, TT was 13 – 17 seconds, fibrinogen was 2.52-3.64 g/L, plasma D-dimer concentration < 0.5µg / ml.  Data were entered into a computerized database structure. Statistical analysis were done using SPSS version 10. The statistical significance of difference in rate of an outcome between 2 groups was assessed by Fisher´s exact significance test, while between more than 2 groups the likelihood test was used. The statistical significance of difference in mean of a continuous variable between 2 groups was assessed by t-test and between more than 2 groups by ANOVA test. The difference in median of an ordered variable (like D-dimer) between 2 groups was assessed by Mann-Whitney test and between more than 2 groups by Kruskal-Wallis test. Spearman´s rank was used to study the strength of association (linear correlation) between D-dimer as an ordered variable and other continuous variables. P value of less than 0.05 was considered statistically significant.

Forty patients were included in the malignancy group and 15 healthy controls were selected.

Thirteen of these patients were admitted to Al- Yermok Teaching Hospital in Baghdad, and the other 27 were recruited from the Hospital of Radiation and Nuclear Medicine in Baghdad. 26 were males (65%) in the malignancy group. The frequency distribution of age and sex in healthy controls and malignant cases are shown in Table 3. There were 26 males (65%) and 14 females (35%), their age ranged from 36 to 73 years for males and from 38 to 70 years for females. The primary sites of malignant disease and their frequency distribution are shown in Figure 1. The most frequent histologic types were adenocarcinoma, 29 patients (72.5%), transitional cell carcinoma, 4 patients (10.0 %), undifferentiated small cell carcinoma, 2 patients (5.0%), non–Hodgkins lymphoma, 2 patients (5.0%) and large cell carcinoma, 1 patient (2.5%). Among the forty cases, 6 (15%) had grade I, 17 (42.5%) had grade II, 12 (30%) had grade III, and 5 (12.5 %) had grade IV tumors, as determined by histology. In about half of the patients (55%, n=22), distant metastasis was present. Thirty-five percent (n=14) patients had slight concurrent bleeding; in 10 of them the bleeding was related to their malignant disease. These were gastrointestinal bleeding from carcinoma of the stomach and colon (n=2), hematuria from carcinoma of the prostate and urinary bladder (n=6) and hemoptysis from the carcinoma of the lung (n=2). While in the other 4 cases, the bleeding was not related to their malignant disease; these were epistaxis, bruising, ecchymosis and vaginal bleeding from the carcinoma of the breast, gall bladder and non – Hodgkin lymphoma. The evidence of thrombosis was encountered in one patient only. All patients were classified according to the results of D-dimer into 2 groups; those with negative D-dimer (considered having no evidence of ICF syndrome), this group consists of 22 patients (55% of cases) and those with positive D-dimer (considered having evidence of ICF syndrome), this group consists of 18 patients (45% of cases). As compared to the healthy controls, PT (p<0.001), aPTT (p<0.001) and TT (p=0.02) were significantly higher in patients with malignancy (Table 1). These variables were also higher in those with evidence of ICF as compared to those without ICF syndrome (Table 2).

Platelet count was significantly higher in patients with malignancy as compared to that in healthy controls (p=0.02) (Table 1). Although lower in ICF patients, the platelet count in patients without ICF was not significantly different from that in patients with ICF (p=0.24). Moreover, the platelet count tended to be lower when there was evidence of ICF of the decompensated type rather than compensated or overcompensated types as depicted in Table 2 (p<0.001). Fibrinogen level was lower in those with ICF (p <0.001) (Table 2). Although not statistically significant, fibrinogen level was slightly lower in decompensated state (2.3±0.3g/L) as compared to compensated (2.6±0.3g/L) and overcompensated ICF states (2.6±0.4g/L) (p=0.62) (Table 2). In this study, 10% (n=4) of cancer patients were diagnosed as decompensated ICF based on prolongation of PT, APTT, low platelet count and/or hyperfibrinogenemia and a positive D-dimer, but 5% (n=2) patients had normal fibrinogen level. Except for D-dimer concentration, the absence, presence of either high-grade tumor or distant metastasis or else the presence of both of these were not associated with a deranged coagulation profile (Table 3).

All patients with thrombocytopenia were those with high grade tumors and not low grade tumors (p=0.009). On the other hand, more subjects with thrombocytosis had evidence of distant metastasis as opposed to the absence of metastasis (n=9 vs. n=2, p=0.04) (Table 4). Thrombocytopenia was observed in five (12.5%) cancer patients; in four of them, the D-dimer was positive, and therefore, they fit the criteria of decompensated ICF, while only one patient had decreased platelet count (30 x 10⁹ /L) with negative D-dimer (Table 4). A total of 16.6% of the patients with positive D-dimer had low fibrinogen level, and as expected, all the patients had prolonged TT, while others had either high fibrinogen level (5.5%) or normal fibrinogen level (77.7%). Thirty-three percent patients with normal fibrinogen level and positive D-dimer also had prolonged TT. The most common coagulation abnormalities in patients with malignancy were elevated D-dimer (45%, n=18), thrombocytosis (27.5%, n=11), prolonged TT (25%, n=10) and hyperfibrinogemia (25%, n=10). Thirty-two patients (80%) had two or more abnormal hemostatic tests (Table 4).

More subjects with adenocarcinoma had evidence of ICF as compared to those with other histologic types but this was not statistically significant (51.9%% vs. 30.8%, p=0.18) (Table 5). On the other hand, ICF was significantly more common in those with high grade tumor than those with low grade tumor (76.5% vs. 21.7%, p=0.001), and in patients showing distant metastasis as compared to those without distant metastasis (72.7% vs. 11.1%, p <0.001) (Table 5). There was a moderately positive linear correlation (r=0.552) between D-dimer concentration and PT, and also a moderately positive linear correlation (r =0.578) between D-dimer and PTT and a strong positive linear correlation (r =0.875) between D-dimer and TT as shown in Table 6. There was a weak negative linear correlation (r= -0.259) between the platelet count and D-dimer concentration (Table 6). There was also a moderately negative linear correlation between D-dimer concentration and fibrinogen level (r= -0.614).

 In this study, 32 patients with malignancy (80%) had two or more abnormal hemostatic tests. The most common abnormalities in patients with malignancy were elevated D-dimer (45%), thrombocytosis (27.5%), prolonged TT (25%) and hyperfibrinogenemia (25%). As compared to the healthy controls, platelet count, PT, aPTT and TT were significantly higher in patients with malignancy. Also, ICF was significantly more common in patients with high grade tumor than those with low grade tumor and in patients with evidence of distant metastasis as compared to those without distant metastasis. Alterations of hemostasis commonly accompany the progression of malignant diseases and every known component of the hemostatic mechanism may be affected by these disease processes. Nearly all patients with an active neoplasm will exhibit at least subtle biochemical changes in hemostasis, and a few of them develop clinically evident thrombosis or hemorrhage [14]. The most common abnormalities in patients with malignancy were elevated D-dimer, thrombocytosis, prolonged TT and hyperfibrinogenemia. These results are comparable with results of a previous study [5]. These abnormalities are consistent with the presence, in cancer patients, of an “overcompensated intravascular coagulation with fibrinolysis (ICF)”. It is hypothesized that low-grade intravascular coagulation with accelerated clotting factor utilization is accompanied by increased synthetic rate for fibrinogen, clotting factors and platelets, resulting in actual increase in their levels in the circulation [15]. In the present study, 80% of the patients had two or more abnormal test results, which is lower than that in a previous study [5] which showed a frequency of 92%. This variation probably can be explained by the differences in type of patients studied because all of the patients in the previous studies were referred to the oncology service as inoperable patients. These studies had a higher percentage (81%) of metastatic disease as compared to our patient sample (55%). Nevertheless, both studies indicated that hemostatic derangement was quite common in cancer patients. The frequency of positive D-dimer test in patients with solid malignant tumors was 45%, which is lower than that in the study conducted by Wilde JT, et al [16], that showed a frequency of 71%. This variation may be explained by the differences in type of patients studied, duration of illnesses and extent of the disease. There was a statistically significant difference in the mean platelet count (p=0.02) between patients with malignancy and the healthy controls. Moreover, in the decompensated state, the mean platelet count was lower as compared to the platelet count in compensated and overcompensated states (p<0.001), and this finding validates a similar finding in a previous study [5].

Thrombocytopenia was observed in five (12.5%) cancer patients. In four of them, the D-dimer was positive, and therefore, they fit the criteria of decompensated ICF, while only one patient had decreased platelet count (30 X 10⁹ /L) with negative D-dimer. This later patient was a case of non-Hodgkin’s lymphoma and had presented with massively enlarged spleen. Splenic pooling and marrow involvement by the disease process are the possible causes for decreased platelet count in this patient. Thrombocytosis was observed in 27.5% (n=11) of the cancer patients, a finding which is consistent with previous studies [4, 5, 17, 18]. In all of these studies, the frequency of thrombocytosis ranged between 25% and 60% [4, 5, 17, 18]. This thrombocytosis, which occurs much more frequently in untreated cancer patients than that in those undergoing treatment for cancer, may be explained by the existence of low-grade DIC, thrombocytolysis and overcompensation. Indeed, thrombopoietic activity has been recovered from the serum of patients with cancer having thrombocytosis [19], suggesting the existence of aberrancy in the usual relationship between platelet count and the production of thrombopoietin in some individuals with cancer. The mean PT and mean aPTT were significantly higher in malignancy cases as compared to healthy controls (p <0.001 and <0.001, respectively), and was also significantly higher in cases with positive D-dimer as compared to those with negative D-dimer (p= 0.003 and 0.007, respectively). Although mean PT and aPTT were higher in decompensated ICF as compared to the compensated and the overcompensated ICF states, the difference were not statistically significant (p=0.82 and 0.5, respectively). These results are in agreement with the result of another study [9]. The mean TT was significantly higher in malignancy cases compared with healthy control (p=0.02), and it was also significantly higher in cases with positive D-dimer compared to those with negative D-dimer (p<0.001), which is consistent with the results of a previous study [5]. The products of fibrin breakdown interfere with fibrin monomers polymerization leading to the prolongation of TT [20]. Mean fibrinogen level was significantly lower in D-dimer positive cases than that in the D-dimer negative malignancy cases (p <0.001). Negative D-dimer in healthy controls in this study and a previous study [16] indicate that plasma D-dimer may be specific for fibrinolysis [10, 21-23]. In this study, 10% (n=4) of cancer patients were diagnosed as decompensated ICF based on prolongation of PT, APTT, low platelet count and/or, hyperfibrinogenemia and positive D-dimer, but half (n=2) of these patients had normal fibrinogen level. This can be explained by an increased fibrinogen production as a result of the underlying disorder and therefore, the increased fibrinogen consumption may not be sufficient to decrease fibrinogen below normal levels early in the decompensated ICF state. As the normal half-life of fibrinogen is approximately 4 days, a 50% or greater decrease in fibrinogen level over one-day is compelling evidence supporting DIC or fibrinolysis, regardless of the final value being normal range [24]. In this study, there was not a significant association between the histological type of tumors and positivity of D-dimer (p=0.18), which is also consistent with a previous study [5], in which there was no relationship between the histological type and the elevation of FDP. However, there was an association between grade of tumors and the positivity of D-dimer. As compared to low grade tumors, more patients with high grade tumors had positive D-dimer (p=0.001). An explanation for this association can be that the tumor cells can directly and/or indirectly activate the coagulation and fibrinolysis systems. It is well-known that most tumor cells can express, on their surface, all proteins necessary for regulating the fibrinolytic pathways. Both, tissue-type plasminogen activator (tPA) and urokinase-type plasminogen activator (uPA), as well as their inhibitors, are expressed in various kinds of tumor cells lines. The expression of uPA seems to be correlated with aggressiveness and histologic grade of tumors as well as clinical progression of different carcinomas [25]. More cases with distant metastasis had positive D-dimer (n=16/22, 72.2%) as compared to those without distant metastasis (n=2/18, 11.1%), (p<0.001). These are consistent with the result of a previous study [26], which showed that the D-dimer level was higher in patients with metastasis than those without metastasis, and that the high plasma D-dimer level is indicative of ongoing fibrinolysis within cancer tissues that occurs during tumor progression. It should be noted that in spite of this evidence for the common occurrence of low-grade ICF in cancer patients, the occurrence of overt DIC, characterized by consumption of platelets and clotting factors with resultant bleeding complication, is uncommon, and DIC of clinical significant occurs in only 9%-15% of patients with cancer [27], which is comparable with findings obtained in the present study in which the percentage of clinical significant DIC was 10%. Limitations of the study include small number of patients and non-application of more specialized tests and recent scoring system that are useful in the diagnosis of this syndrome such as measurement of soluble fibrin monomers and the sensitive assays that can measure the generation of thrombin fragment F1+2 or thrombin- antithrombin complex.

 Disorders of hemostasis are present in a majority of patients with solid tumor malignancy. In a proper clinical setting, supported by abnormal hemostatic test results, the ICF syndrome should be suspected. Furthermore, precautionary measures can be taken to avoid its complication. Local or systemic activation of blood coagulation can be produced by tumor products and favors tumor spread, while interruption of blood coagulation reactions, in general, favors the host and impairs metastasis [27].

1. Loreto MF, Martinis MD, Corsi MP, Modesti M, et al. Coagulation and cancer, Pathology Oncology Research. 2000; 6:302-312.
2. Soong BCF, Miller SO: Coagulation disorders in cancer, III fibrinolysis and inhibitor. Cancer 1970; 25: 4: 867 – 873.
3. Ambrus JL, Ambrus CM, Mink IB, pickren JW, et al: Causes of death in cancer patients. J Med 1975; 6: 61 – 64.
4. Hagedorn AB, Bowie Ejw, Elveback LR, Owen CA, et al: Coagulation abnormalities in patients with inoperable lung cancer. Mayo Cli Proc 1974; 49: 649 – 653.
5. Sun NC, MCAfee WM, Hum GJ, Weiner JM, et al: Haemostatic abnormalities in malignancy: A prospective study in one hundred eight patients, part 1. Am J Clin Path 1979; 71: 10 – 16.
6. Gordon SG, crosses BA: Factor X- activity cystic protease from malignant tissue. J Clin Invest 1981; 67: 1665 – 1671.
7. Hagedorn AB, Bowie Ejw, Elveback LR, Owen CA, et al: Coagulation abnormalities in patients with inoperable lung cancer. Mayo Clin Proc 1974; 49: 649 – 653.
8. Sun NC, MCAfee WM, Hum GJ, Weiner JM, et al: Haemostatic abnormalities in malignancy: A prospective study In one hundred eight patient, parts 1. Am J Clin Path 1979; 71: 10 – 16.
9. Seligsohn URI: Disseminated intravascular coagulation in: Beutler E, Lichtman MA, coller BS (ed).Williams Hematology, 5th ed, MC Graw – Hill publishing company, USA, 1995; 1497 – 1516.
10. Greenberg CS, Devine DV, and McCrae KM: Measurment of plasma D- dimer levels with the use of a monoclonal antibody coupled to latex beads. Am J Cli Pathol 1987; 1: 94-100.
11. Taylor FB, Toh CH, Hoots WK, Wada H, Levi M, et al: Definition and scoring system for Disseminated Intravascular coagulation. Thromb Haemost 2001; 86: 1327 – 1330.
12. Okajima K, Sakamoto Y, Uchiba M: Heterogeneity in the incidence and clinical manifestations of disseminated intravascular coagulation: Study of 204 cases. Am J Haematol 2000; 65: 215–222.
13. KobayashN, Maegawa T, Takada M, Tanaka H, Gonmoie H, et al: Criteria for diagnosis of DIC based on the analysis of clinical and laboratory finding in 345 DIC patients collected by the research committee on DIC in Japan. Bibl Haematol 1987; 49:265 – 275.
14. Goldsmith GH: Haemostatic changes in patients with malignancy. Int J Haematol 2001; 73(2): 151 – 156.
15. Frederick R, Richard L: Activation of blood coagulation in cancer: Trousseau’s syndrome revisited. Blood 1983;62: 1: 14 –31.

16. Wilde JT, kitchens, Kinseys, et al: Plasma D – dimer levels and their relationship to serum fibrinogen \ fibrindegradation products in hypercoagulable state. Br J Haematol 1989; 71: 65 – 70.

17. Levin J, Conely cL: Thrombocytosis associated with malignant Arch Inter Med 1964; 114: 497 – 500.
18. Davis RB, Theologies A, Kennedy BJ: Comparative studies of blood coagulation and platelet aggregation in patients with cancer and non – malignant disease. Ann Int Med 1969; 71: 67 – 80.
19. Saikh BS, Bullard V: Thrombopoietic activity from sera of cancer patients. Blood 1977; 50(suppl 1): 252.
20. Zakim D, Boyer D Thomas: Hepatology Test Book Of Liver Disease, 3rd ed, W.B Saunders Company, London, 2001; 550-580.
21. Andrew F, Huatiuk W: Circulating D – dimer in patients with Chest 2002; 117: 1012 – 1016.
22. Elms MJ, Bundesen PG, Rylatt DS, et al: Rapid detection of cross-linked fibrin degradation products in plasma using monoclonal antibody coated latex particles. Am J Clin Pathol 1986; 3: 360 – 364.
23. Elms MJ, Bunce IH, Bundesen PG, et al: Measurement of cross linked fibrin degradation products an immunoassay using a monoclonal antibodies. Thromb Haemost 1983; 50: 591 – 594.
24. Trofatter KF, Howell ML, Greeb berg et al: Use of the fibrin D- dimer in screening for coagulation abnormalities in preeclampsia. Obstet Gynecol 1989; 3: 435 – 439
25. Hajjar KA, Esmon NL, Marcus AJ :Vascular function in: Bulter E, Lightman MA , Culler BS,Kipps TJ.Williams Haematology 6th ed, New York. Mc Graw- Hill Company 2001; chapter 114:1451-1469.
26. Kin HK, song KS, Kyung Sb, Lee KR, kany YH, Lee YH, Lee ES, et al: Comparison of plasma D – dimer and thrombus precursor protein in patients with operable breast cancer as a potential predictor of lymph node metastasis. Blood Coagulation and Fibrinolysis 2004; 15: 1- 13 (Abstract).
27. Frederick R, Richard L: Activation of blood coagulation in cancer: Trousseau’s syndrome revisited. Blood 1983; 62: 1: 14 –31.