Detection of a Hypercoagulable State in Patients With Pancreatic Cancer Using a Clot Waveform Analysis

1. Introduction

Thrombotic complications such as venous thromboembolism (VTE), acute cerebral infarction (ACI) and acute coronary syndrome (ACS) have often been reported and studied as cancer-associated thrombosis (CAT) by many physicians, and researchers [1,2,3]. The incidence of VTE is <0.1 % in the general population and it was reported that approximately 20% of the overall VTE incidence might be caused by active cancer [4]. CAT including among ACS, ACI and VTE, is considered a main cause of noncancer death in patients with cancer [5]. VTE is one of the most important thrombotic complications in patients with cancer [6,7], and ACS and ACI are arterial thromboses that are mainly caused by platelet activation and atherosclerosis [8,9]. Trousseau syndrome [10,11] is a well-known thrombotic syndrome in cancer patients. CAT, including Trousseau syndrome is associated with various thrombotic diseases and has various factors, including leukocytosis, thrombocytosis, physical factors and increased coagulation factors, which have been proposed as risk factors for CAT [1].

Hypercoagulability including increased levels of coagulation factors, such as tissue factor (TF) and clotting factor VIII (FVIII), activated coagulation factors and TF-positive microparticles was proposed as a risk factor for CAT [1]. Activation of the coagulation cascade and platelets which is related with tumor growth and metastasis, may cause CAT [12,13]. Therefore, an early diagnosis of hypercoagulability and prophylaxis for CAT may be important in management of cancer [10].

Although increased D-dimer or soluble fibrin (SF) levels are useful for detecting thrombosis in patients with CAT [14,15,16], routine coagulation tests, such as activation partial thromboplastin time (APTT) and prothrombin time (PT) cannot detect the hypercoagulable state [17]. A clot waveform analysis (CWA) of APTT (CWA-APTT) has been developed to analyze to evaluate hemostatic abnormalities in patients with bleeding tendency and to monitor anticoagulant therapy [17,18,19]. Furthermore, the use of a small amount of TF induced FIX activation (sTF/FIXa) (CWA-sTF/FIXa) assay can evaluate the hemostatic abnormalities including platelet abnormalities [17]. Hypercoagulability is considered to be caused by a thrombin burst [20,21] which is evaluated by thrombin time using a small amount of thrombin with a CWA (CWA-TT) [22].

In the present study, hypercoagulability and complications of thrombosis in patients with malignant neoplasms including pancreatic cancer were investigated using a CWA-APTT, CWA-sTF/FIXa and CWA-TT in comparison to patients without cancer or those with benign pancreatic diseases.

2. Materials and Methods

The study population included 138 patients with cancer and 66 patients without cancer who were managed at Mie Prefectural General Medical Center from August 21, 2020 to December 31, 2022. Cancer types were pancreatic cancer (n=30), uterine cancer (n=4), prostate cancer (n=7), hepatocellular carcinoma (n=15), malignant lymphoma (n=9), leukemia and myelodysplastic syndrome (leukemia/MDS) (n=11), lung cancer (n=11), breast cancer (n=5), biliary tract cancer (n=11), colon cancer (n=10), stomach cancer (n=12), esophageal cancer (n=5) and other cancer (n=8; origin unknown, n=3; bladder cancer, n=2; kidney cancer, n=2 and ovarian cancer, n=1) (Table 1). Noncancer patients included those with chronic hepatitis (n=31), anemia (n=9), bone marrow proliferative disease (n=6), pregnancy loss (n=3), fatty liver (n=3), carrier of hepatitis B virus (n=4), alcoholic liver damage (n=2), hepatic hemangioma (n=2), pancreatic cyst (n=3), benign pancreatic tumor (n=2) and pancreatitis (n=1). CWA was also performed in 20 healthy volunteers as controls. Disseminated intravascular coagulation (DIC) was diagnosed using the Japanese Ministry of Health Labor and Welfare criteria for DIC [23]. Acute cerebral infarction was diagnosed using computed tomography or magnetic resonance imaging, acute coronary syndrome was diagnosed using coronary angiography, electrocardiography and was based on elevated troponin levels, venous thromboembolism was diagnosed using computed tomography or venous ultrasound [24]. The stage of pancreatic cancer was determined in accordance with guidelines [25].

The study protocol (2019-K9) was approved by the Human Ethics Review Committee of Mie Prefectural General Medical Center, and informed consent was obtained from each participant. This study was carried out in accordance with the principles of the Declaration of Helsinki.

The CWA-APTT was performed using APTT-SP^®, which uses silica as an activator of FXII and synthetic phospholipids (Instrumentation Laboratory; Bedford, MA, USA) and platelet-poor plasma (PPP), with an ACL-TOP^® system (Instrumentation Laboratory) as previously reported [17,26]. The CWA- sTF/FIXa assay was performed using platelet-rich plasma (PRP) and 2,000-fold diluted HemosIL RecombiPlasTin 2G (TF concentration <0.1 pg/ml; Instrumentation Laboratory) [27]. Three curves are shown on the monitor of this system [17], one shows the changes in the absorbance observed while measuring the APTT, corresponding to the fibrin formation curve (FFC), the second (first derivative peak 1^st DP), corresponds to the coagulation velocity and the third (second derivative peak 2^nd DP) corresponds to the coagulation acceleration. The respective height and time of the 1^st DP, 2^nd DP and fibrin formation are called the 1^st DP height (1^st DPH) and 1^st DP time (1^st DPT), 2^nd DPH and 2^nd DPT, and fibrin formation height (FFH) and fibrin formation time (FFT), respectively (Figure 1). PRP was prepared by centrifugation at 900 rpm for 15 minutes (platelet count, 40 x 10¹⁰ /L), and PPP was prepared by centrifugation at 3,000 rpm for 15 minutes (platelet count, <0.5 x 10¹⁰ /L) [28].

The CWA-TT was measured using 0.5 IU thrombin (Thrombin 500 units Mochida Pharmaceutical CO., LTD, Tokyo, Japan) with an ACL-TOP^® system (Instrumentation Laboratory) [17]. Three types of curves are shown on this system monitor [22]. The 1^st DP shows two peaks and the second peak height of the 1^ST DP (1^st DPH-2) reflects the thrombin burst.

Statistic alanalyses

The data are expressed as the median (range). The significance of differences between groups was examined using the Mann-Whitney U-test. P values of <0.05 were considered to indicate a statistical significance. All statistical analyses were performed using the Stat-Flex software program (version 6; Artec Co., Ltd, Osaka, Japan).

3. Results

The frequency of thrombotic complications was high in cases of pancreatic cancer (36.7 %), prostate cancer (43.0 %), malignant lymphoma (44.4 %) and others (37.5 %) (Table 1). Thrombotic complications were seen in 18 patients with VTE (pancreatic cancer 8 and other cancer 10), 6 patients with acute cerebral infarction (pancreatic cancer 1 and other cancer 5), 4 patients with DIC (pancreatic cancer 1 and other cancer 3), one patient with acute coronary syndrome (pancreatic cancer) and one patient with thrombotic microangiopathy (other cancer). The frequency of thrombosis tended to be higher in patients with pancreatic cancer than in those with other cancers (19.4 %), although this difference was not significant. The thrombotic complications in patients with other cancers was gradually increased as the stage of cancer progressed; however, a high frequency of thrombotic complications (60%) was observed in stage I pancreatic cancer (Table 2).

Figure 1 shows the results of the CWA-APTT in a patient with pancreatic cancer (a), esophageal cancer (b), chronic hepatitis (c) and pancreas cyst (d). The peak heights of the FFC, 1^st DP and 2^nd DP were highest in the patient with pancreatic cancer, followed by the patient with esophageal cancer. Figure 2. Shows the results of statistical analysis for the CWA-APTT in patients with cancer, patients without cancer and healthy volunteers (upper column), cancer patients with and without pancreatic cancer, and patients with benign pancreatic diseases (lower column). The peak times of the CWA-APTT in patients with cancer, (median, 2^nd DPT 33.9 sec; 1^st DPT 36.9 sec; FFC 39.5 sec) were significantly longer than those in patients without cancer (median, 2^nd DPT 31.9 sec, p <0.01; 1^st DPT 34.8 sec, p <0.001; FFC 36.4 sec, p <0.001) and healthy volunteers (median, 2^nd DPT 31.7 ^$1, two with VTE and one with acute coronary syndrome; ^$3, two with VTE; ^$4, four with VTE, one with DIC and one with ACI; ^#1, two with VTE and one with ACI; ^#2, two with VTE and one with ACI; ^#3, two with VTE, one with DIC and one with ACI; ^#4, four with VTE, two with ACI, two with DIC and one with thrombotic microangiopathy; VTE, venous thromboembolism; DIC, disseminated intravascular coagulation; ACI, acute cerebral infarction.

CWA, clot waveform analysis; APTT, activated partial thromboplastin time; navy line, fibrin formation curve; pink line, 1^st derivative curve (velocity); light blue, 2^nd derivative curve (acceleration); solid line, patient; dotted line, healthy volunteer

	Cancers	Without cancer	PCs	Without PC	BPDs	HVs
2nd DPT (sec)	33.9 (31.1-38.1)	31.9 (29.7-34.4)	34.8 (32.6-38.0)	33.6 (30.8-38.1)	29.1 (28.3-33.9)	31.7 (31.1-32.9)
2nd DPH (mmAbsorbance)	953 (792-1143)	761 (661-866)	1094 (850-1181)	940 (786-1110)	898 (633-1007)	734 (609-800)
1st DPT (sec)	36.9 (34.0-41.7)	34.8 (32.5-37.8)	38.4 (36.2-40.6)	36.5 (33.6-42.0)	31.5 (30.3-37.4)	34.3 (33.8-35.9)
1st DPH (mmAbsorbance)	345 (258-434)	255 (212-292)	353 (264-436)	331 (257-406)	325 (263-459)	235 (209-270)
FFC (sec)	39.5 (36.4-44.8)	36.4 (34.5-39.7)	38.9 (36.1-44.8)	40.6 (37.9-45.6)	33.7 (33.0-39.1)	36.0 (35.6-38.2)
FFH (mmAbsorbance)	285 (225-411)	201 (185-220)	355 (263-426)	274 (221-361)	211 (202-225)	187 (170-206)

The peak heights in patients with cancer (median, 2^nd DPH 953 mm absorbance; 1^st DPH 345 mm absorbance; FFC 285 mm absorbance) were also significantly higher than those in patients without cancer (median, 2^nd DPH 761 mm absorbance, p <0.001; 1^st DPH 255 mm absorbance, p <0.001; FFC 201 mm absorbance, p <0.001) and healthy volunteers (median, 2^nd DPH 734 mm absorbance, p <0.001; 1^st DPH 235 mm absorbance, p <0.001; FFC 187 mm absorbance, p <0.001). There were no significant differences in the peak times or heights of the CWA-APTT between patients without cancer and healthy volunteers. The peak times of the CWA-APTT were significantly longer in patients with pancreatic cancer than patients with benign pancreatic diseases (median, 2^nd DPT 34.8 sec vs. 29.1 sec, p <0.05; 1^st DPT 38.4 sec vs. 31.5 sec, p <0.05; FFC 38.9 sec vs. 33.7 sec, p <0.05). There were no significant differences in the peak times between patients with pancreatic cancer and those with other cancers. The 1^st DPH of CWA-APTT was significantly higher in patients with pancreatic cancer (median, 353 mm absorbance) than in patients with other cancers (median, 331 mm absorbance, p <0.05) and patients with benign pancreatic diseases (median, 325 mm absorbance, p <0.01). The FFC was significantly higher in patients with pancreatic cancer (median, 355 mm absorbance) than patients with benign pancreatic diseases (median, 211 mm absorbance, p <0.01).

Regarding the CWA-sTF/FIXa in PRP without phospholipids (Figure 3), the peak times of the CWA-sTF/FIXa were shorter in patients without cancer than in patients with CWA, clot waveform analysis; sTF/FIXa, small amount of tissue factor induced FIX activation; navy line, fibrin formation curve; 1^st derivative curve (velocity); light blue, 2^nd derivative curve (acceleration); solid line, patient; dotted line, healthy volunteer cancers and healthy volunteers. The peak heights of FFC, 1^st DP and 2^nd DP were highest in patient with pancreatic cancer followed by patient with esophageal cancer. In the statistical analysis (Figure 4), the peak times of the CWA-sTF/FIXa were shorter in patients without cancer (median, 2^nd DPT 70.2 sec; 1^st DPT 85.0 sec; FFC 86.2 sec) than in patients with cancer (median, 2^nd DPT 78.3 sec, p <0.001; 1^st DPT 95.9 sec, p<0.001 and FFC 97.2 sec, p <0.001) and healthy volunteers (median, 2^nd DPT 76.9 sec, p <0.01; 1^st DPT 95.3 sec, p<0.001 and FFC 94.6 sec, p <0.01). The 2^nd DPT values of patients with pancreatic cancer (median, 78.1 sec, p <0.05) and with other cancer (median, 78.3 sec, p <0.05) were significantly longer than those of patients with benign pancreatic diseases (median, 59.7 sec). There were no significant differences in peak times among patients with pancreatic cancer, other cancers and benign pancreatic diseases). The 2^nd DPH (median, 47.1 mm absorbance vs. 36.9 mm absorbance, p <0.01), 1^st DPH (median 95.9 mm absorbance vs. 85.0 mm absorbance, p <0.001) and FFH (median, 384 mm absorbance vs. 324 mm absorbance, p <0.001) were significantly higher in patients with cancer than in those without cancer. The 1^st DPH was significantly higher in patients with pancreatic cancer (median, 127 mm absorbance) than in patients with other cancers (median, 96.1 mm absorbance, p <0.05) and patients with benign pancreatic diseases (median, 74.7 mm absorbance, p < 0.01). The FFH (median, 463 mm absorbance vs. 268 mm absorbance, p < 0.01) was significantly higher in patients with pancreatic cancer than patients with benign pancreatic diseases. There was no significant difference in the 2^nd DPH or FFH between patients with pancreatic cancer and patients with other cancers.

	Cancers	Without cancer	PCs	Without PC	BPDs	HVs
2ndDPT (sec)	78.3 (67.5-87.0)	70.2 (66.1-75.4)	78.1 (68.9-83.6)	78.3 (67.1-86.9)	59.7 (53.5-76.8)	76.9 (70.1-84.0)
2ndDPH (mmAbsorbance)	47.1 (29.4-68.2)	36.9 (31.5-43.1)	61.5 (36.3-70.3)	44.8 (28.3-67.5)	40.3 (22.7-58.8)	37.4 (31.0-48.1)
1stDPT (sec)	95.9 (84.0-110.3)	85.0 (81.2-90.9)	93.8 (84.5-108.7)	96.0 (82.8-110.7)	83.6 (65.5-102.3)	95.3 (89.8-106.4)
1stDPH (mmAbsorbance)	104 (74.9-155)	75.6 (65.8-83.7)	127 (84.8—165.0)	96.1 (72.1-139.6)	74.7 (55.5-84.7)	67.8 (58.2-74.5)
FFC (sec)	97.2 (86.0-110.4)	86.2 (82.2-91.6)	95.7 (86.1-109.1)	97.4 (84.8-111.2)	95.4 (75.2-101.4)	94.6 (89.4-106.9)
FFH (mmAbsorbance)	384 (306-526)	324 (267-344)	463 (354-565)	374 (301-482)	268 (256-275)	341 (278-378)

Regarding CWA-TT, the peak heights were significantly higher in PRP than in PPP. The second 1^st DPH was extremely high in patients with pancreatic cancer (Figure 5). In the statistical analysis, the 2^nd DPT, 1^st DPT-1 and FFC were significantly shorter in cancer patients (median, 34.6 sec, p< 0.001; 40.5 sec, p < 0.001; and 108 sec, p < 0.001, respectively) and patients without cancers (median, 34.7 sec, p <0.001; 41.0 sec, p <0.001 and 129 sec, p <0.001, respectively) than in healthy volunteers (median, 42.7 sec; 49.9 sec and 138 sec, respectively). Only the 1^st DPT-2 was significantly shorter in patients with cancer (median, 116 sec) than in patients without cancer (median, 145 sec, p <0.001) and healthy volunteers (median, 133 sec, p < 0.01) (Figure 6). There were no significant differences in peak times among patients with pancreatic cancer, other cancers or benign pancreatic diseases. The 2^nd DPH (median, 241 mm absorbance vs. 102 mm absorbance, p < 0.001), 1^st DPH-1 (median, 146 mm absorbance vs. 82.5 mm absorbance, p <0.001), 1^st DPH-2 (median, 82.6 mm absorbance vs. 27.9 mm absorbance, p <0.001) and FFH (median, 922 mm absorbance vs. 695 mm absorbance, p <0.001) were significantly higher in patients with cancer than in healthy volunteers. The 2^nd DPH, 1^st DPH-2 and FFH were significantly higher in patients with cancer than in patients without cancer (median, 141 mm absorbance, p < 0.001; 51.8 mm absorbance, p <0.001 and 911 mm absorbance, p <0.01). The 2^nd DPH and 1^st DPH-1 were significantly lower in patients with benign pancreatic diseases than in patients with pancreatic cancer (median, 157 mm absorbance vs. 261 mm absorbance, p <0.05 and 98.5 mm absorbance vs. 176 mm absorbance, p <0.05, respectively) or other cancers (median, 237 mm absorbance, p <0.05 and 141 mm absorbance, p <0.05, respectively). There were no significant differences in the 1^st DPT-2 or FFH among patients with pancreatic cancer, other cancers or benign pancreatic diseases.

4. Discussion

CAT is a condition whose relevance has been recognized frequently both among physicians who deal with VTE and oncologists [29]. The annual incidence of VTE is 0.5% in cancer patients compared to <0.1% in the general population and active cancer accounts for 20% of the overall incidence of VTE [29]. Several mechanisms underlying the prothrombotic state in these patients have been proposed. Increased levels of clotting factors, leukocytes, platelets, and TF-positive microparticles are all potential factors that alone or in combination increase CAT [30]. After activation of coagulation system, fibrin- related markers, such as D-dimer, SF and FDP are increased, and these biomarkers are used for the diagnosis and exclusion of VTE [31,32]. The platelet activation receptor C-type lectin-like receptor 2 (CLEC-2), and its endogenous ligand podoplanin have attracted the attention of many researchers [33].

In this study, 24.6% of patients with thrombotic diseases had malignant neoplasms, indicating that these cancer patients were in a hypercoagulable state. In particular, 36.7% of pancreatic cancer patients were associated with thrombosis, suggesting that pancreatic cancer patients are markedly hypercoagulable and have a high risk for thrombosis. Patients with pancreatic cancer have a very high risk of both venous and arterial thrombosis compared to those with other cancers, caused by their tumor-driven hypercoagulable state [34]. The present study also showed the significant prolongation of peak times of CWA-APTT which is equal to routine APTT, in cancer patients including those with pancreatic cancer. Although these findings may prompt the misunderstanding that hypocoagulability exists in cancer patients, the markedly increased peak heights of the CWA-APTT and CWA- sTF/FIXa actually suggest a hypercoagulable state in patients with cancer including those with pancreatic cancer. The APTT using massive phospholipids and substances that activate the contact system does not reflect physiological coagulation, whereas the sTF/FIXa using PRP instead of massive phospholipids, does reflect physiological coagulation.

	Cancers	Without cancer	PCs	Without PC	BPDs	HVs
2ndDPT (sec)	34.6 (31.6-39.8)	34.7 (31.3-39.9)	35.8 (32.0-41.1)	34.0 (30.9-39.6)	35.9 (31.2-41.1)	42.7 (40.3-45.8)
2ndDPH (mmAbsorbance)	241 (148-352)	141 (112-185)	261 (167-387)	237 (150-351)	157 (122-160)	102 (83.7-113.4)
1stDPT-1 (sec)	40.5 (36.3-46.0)	41.0 (36.9-46.3)	41.7 (37.1-47.0)	39.9 (35.6-46.0)	42.6 (36.5-49.3)	49.9 (47.9-52.9)
1stDPH-1 (mmAbsorbance)	146 (109-205)	92.4 (81.0-110)	176 (118-218)	141 (106-204)	98.5 (89.2-117)	82.5 (74.8-94.9)
1stDPT-2 (sec)	116 (97.3-133.7)	145 (121-161)	112 (97.3-133.2)	117 (97.4-133.6)	122 (NC)	133 (127-141)
1stDPH-2 (mmAbsorbance)	82.6 (60.3-106.3)	51.8 (42.6-68.0)	81.4 (64.5-122.2)	83.6 (58.4-104.8)	56.5 (NC)	27.9 (23.3-32.9)
FFC (sec)	108 (90.7-124.5)	129 (114-147)	109 (95.1-128.8)	107 (88.9-124.3)	113 (98.1-126.6)	138 (128-147)
FFH (mmAbsorbance)	922 (822-1055)	850 (779-933)	975 (827-1113)	911 (821-1032)	796 (729-869)	695 (618-729)

The mechanism underlying hypercoagulability in pancreatic cancer [35,36,37] involves increased procoagulant factors, including fibrinogen [38], FVIII [39] and TF [40,41] and reduced anticoagulant factors [36,42]. As fibrinogen is measured by clotting assay, the fibrinogen activity may be higher than the fibrinogen antigen in the hypercoagulable state. Markedly high peak heights of the sTF/FIXa using small amounts of TF suggests that intrinsic TF released cancer cells exist in plasma. Thrombin causes an elevation in FVIII activity and the activation cycle from thrombin to FXIa, called the thrombin burst [22]. Elevated peak heights, especially the second peak of the 1^st DPH (1^st DPH-2) of CWA-TT suggest that the thrombin burst worsens hypercoagulability in cancer patients including those with pancreatic cancer. Platelet aggregation induced by tumor cells is a key contributor to the prothrombotic state in pancreatic cancer. Pancreatic cancer cell lines induce platelet aggregation in vitro via a thrombin-dependent mechanism [43]. In the present study, the CWA-sTF/FIXa assay in PRP showed hypercoagulability and the CWA-TT showed that the peak heights were significantly higher in PRP than in PPP, suggesting that platelets may play an important role in hypercoagulability in cancer patients. Human pancreatic tumors and cell lines have been shown to express plasminogen activator inhibitor type 1 (PAI-1), a key inhibitor of fibrinolysis [44], suggesting that the evaluation of fibrinolysis is important. However, CAT with hyperfibrinolysis may cause DIC in patients with advanced stomach cancer [45], whereas DIC is rare in patients with pancreatic cancer. TF borne by circulating microparticles released from TF-expressing pancreatic cancer cells might contribute to thrombosis formation [46]. Elevated cytokines [47] and neutrophil extracellular traps [48] may cause inflammatory thrombosis. These reports suggest that an evaluation of white blood cells including neutrophils and monocytes might be important.

5. Conclusions

Elevated peak heights of CWA-APTT, CWA-sTF/FIXa and CWA-TT indicated marked hypercoagulability induced by TF released from cancer cells, platelet activation and thrombin burst in patients with malignant neoplasms including pancreatic cancer.