Abstract
Portal vein thrombosis (PVT) is a serious complication after hepatobiliary-pancreatic surgery. Portal vein thrombosis often develops in patients with liver cirrhosis (LC) postoperatively, although they have low platelet counts. Platelet activation is one of the causes of thrombosis formation, and soluble form of glycoprotein VI (sGPVI) has received attention as a platelet activation marker. We had prospectively enrolled the 81 consecutive patients who underwent splenectomy (Sx) and/or hepatectomy: these patients were divided as Sx (n = 38) and hepatectomy (Hx, n = 46) groups. The 3 patients who underwent both procedures were added to both groups. Each group was subdivided into patients with non-LC and LC: non-LC-Sx (n = 22) and LC-Sx (n = 16), non-LC-Hx (n = 40) and LC-Hx (n = 6). The presence of PVT was diagnosed by using enhanced computed tomography (CT) scan. Platelet counts were significantly lower in LC-Sx than in non-LC-Sx, and incidence of PVT was significantly higher in LC-Sx than in non-LC-Sx (68.8% vs 31.8%, P = .024). Soluble form of glycoprotein VI /platelet ratios on preoperative day and postoperative day 1 were significantly higher in LC-Sx than in non-LC-Sx. Incidence of PVT was significantly higher in LC-Hx than in non-LC-Hx (50.0% vs 7.5%, P < .01). Soluble form of glycoprotein VI /platelet ratios were significantly higher in LC-Hx before and after Hx, compared to non-LC-Hx. Patients with LC stay in hypercoagulable state together with platelet activation before and after surgery. Under this circumstance, alteration of portal venous blood flow after Sx or Hx is likely to cause PVT in patients with LC.
Keywords: portal vein thrombosis, glycoprotein VI, hepatobiliary-pancreatic surgery
Introduction
Portal vein thrombosis (PVT) is a common complication after the hepatobiliary-pancreatic surgery, such as splenectomy (Sx) and hepatectomy (Hx); and its incidence was reported to be 2.6% to 16.9%1,2 after Sx, 2.1% to 9.1%3,4 after Hx, respectively. According to the definition of PVT and frequency of computed tomography (CT) scan after operation, however, the incidences of PVT markedly differ. In the previous study in which postoperative enhanced CT scan was performed to evaluate PVT including thrombosis in splenic vein in 33 consecutive patients with Sx, the incidence of PVT was reported to be very high, being 52%.5 Portal vein thrombosis causes serious liver dysfunction, hepatic failure, and exacerbation of varices and leads to a poor outcome.6 It was previously reported that the development of PVT after the hepatobiliary-pancreatic surgery frequently occurred in the patient with liver cirrhosis (LC)7 although they had low platelet counts. It is generally accepted that decreased portal venous flow is the primary factor leading to the development of PVT in patients with LC.8 However, as for the cause of PVT, not only decreased portal venous flow but also the coagulation disorder are regarded as an important factor.9–11
Major determinants of the thrombogenicity of the blood vessel wall are collagens of the subendothelium. The interaction of platelets with collagen involves firstly adhesion and, subsequently, activation leading to second phase adhesion, secretion, and ultimately aggregation. Platelet activation is a vital step in thrombosis formation. Especially, in the process of platelet activation, glycoprotein VI has become to be known as a major collagen receptor.12 Glycoprotein VI is a transmembrane glycoprotein expressing on the platelet and is a collagen receptor belonging to immunoglobulin superfamily. Glycoprotein VI plays an important role not only in the adhesion to the exposed collagen in vascular endothelial injured part but also in signal transmission of the platelet activation by the collagen. Glycoprotein VI is constitutively associated and expressed with the Fc receptor γ-chain (FcRγ), and FcRγ has an immunoreceptor tyrosine-based activation motif-bearing.13,14 When a ligand is connected to the GPVI receptor, the intracellular domain of the FcRγ subunit which forms dimer with GPVI receptor is phosphorylated, and an activated signal is transmitted. Upon platelet activation, the platelet surface GPVI is cleaved off by protease, such as a disintegrin and metalloproteinase domain–containing protein 10 , releasing the soluble-form GPVI (sGPVI).15–17Soluble-form GPVI is regarded as a useful biomarker to evaluate platelet activation because of the following features: (1) its expression is platelet-specific, (2) its plasma levels are elevated in disorders marked by platelet activation, (3) its generation is well controlled and depends on platelet activation.18
According to the previous literatures, sGPVI levels are elevated in the diseases caused by the thrombus and coagulation disorder, such as acute coronary syndrome (ACS),19 stroke,20 disseminated intravascular coagulation,21 and deep vein thrombosis (DVT).22 Recently, Egan et al reported that sGPVI levels were significantly higher in patients with LC than those in healthy patients.23 To the best of our knowledge, there have been no reports on the relationship between sGPVI levels and PVT.
In the present study, based on the hypothesis that activation of the platelet is one of the major causes of PVT development especially in patients with LC, we prospectively measured sGPVI levels in the patients undergoing Hx and Sx in an attempt to clarify the relationship between sGPVI levels and development of PVT.
Materials and Methods
Between February 2012 and February 2014, at Mie University Hospital, 81 consecutive patients who underwent Sx and/or Hx had been prospectively enrolled in this study, because these procedures are known to have high incidence of PVT postoperatively.
In the 81 patients, operative procedure was Sx alone in 22, DP with Sx in 13, Hx alone in 43, and Sx and Hx in 3. The primary diseases for Sx alone were pancytopenia due to LC in 7, chronic hepatitis in 5, hepatocellular carcinoma (HCC) in 2, splenic malignant lymphoma in 2, splenic metastasis of other organs cancer in 2, idiopathic thrombocytopenic purpura in 2, splenic abscess in 1, and idiopathic portal hypertension in 1. Those for DP with Sx were pancreatic cancer in 9, intraductal papillary mucinous neoplasm in 3, and pancreatic neuroendocrine tumor in 1. Those for Hx alone were perihilar cholangiocarcinoma in 17, donor of living donor liver transplantation in 11, HCC in 7, cholangiocellular carcinoma in 3, hemangioma in 2, mucinous cyst neoplasm in 1, peribiliary cyst in 1, and intraductal papillary neoplasm of bile duct in 1. Those for Sx and Hx were HCC and pancytopenia due to chronic hepatitis in 3.
These patients were divided into the 2 groups according to the operative procedure: Sx(n = 38) and Hx(n = 46) groups (Figure 1). The 3 patients who underwent both Sx and Hx were included in both groups. Since the presence of LC is considered to influence the development of postoperative PVT, each group was subdivided into patients with non-LC and LC: non-LC-Sx (n = 22) and LC-Sx (n = 16) in Sx group, non-LC-Hx (n = 40) and LC-Hx (n = 6) in Hx group.
Figure 1.
Methods for patient grouping according to operative procedures and presence of liver cirrhosis (LC). Hx indicates hepatectomy; Sx, splenectomy.
In the 2 groups, we had prospectively measured the following parameters in blood on the day before operation and on postoperative day (POD) 1, 3, 5, 7, 10, and 14, to compare between patients with non-LC and LC: platelet counts, and plasma levels of sGPVI, von Willebrand factor (VWF), VWF propeptide (VWFpp), and a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 (ADAMTS13).
The platelet counts were measured using the fully automated hematology analyzer XE-2100 (Sysmex, Kobe, Japan). The plasma level of sGPVI was quantified using a sandwich enzyme-linked immunosorbent assay, which consisted of 2 mouse anti-GPVI monoclonal antibodies, F1232-7-1 and F1232-10-2 able to recognize the extracellular domain 1 (D1) N-terminal loop and extracellular domain D2 loop of GPVI, respectively.24,25 The VWF and VWFpp levels were measured with a VWF & Propeptide Assay kit (GTI Diagnostics, Waukesha, Wisconsin).26 The ADAMTS13 was measured using FRETS-VWF73, which was chemically synthesized by the Peptide Institute, Inc (Osaka, Japan) according to the method of Kokame et al.27
The present study was approved by the ethical committee of Mie University Hospital in accordance with the ethical standards established in the Declaration of Helsinki (No. 2317).
Diagnostic Criteria of LC and PVT
Liver cirrhosis was diagnosed on the basis of combination of clinical features, blood profile, and radiological imaging according to Nangliya et al.28 Clinical features included symptoms of portal hypertension, such as ascites and/or gastrointestinal varices. Blood profile included evidence of thrombocytopenia (platelet counts less than 100 × 103μL) and/or coagulopathy. Radiological features of CT demonstrated a small shrunken liver and/or irregular liver surface with or without splenomegaly and gastrointestinal varices. The presence of postoperative PVT was diagnosed by using enhanced CT scan. Portal vein thrombosis is defined as the thrombus in splenic vein, thrombus continued from splenic vein to portal vein, and thrombus in portal vein or superior mesenteric vein.
Statistical Analysis
Quantitative variables were presented as the medians and range. Categorical variables were expressed as numbers and percentages. Mann-Whitney U test was used to compare quantitative variables. The χ2 test was used to compare categorical variables. A P value of less than .05 was considered statistically significant. All analyses were performed using the IBM SPSS statistics 24 (IBM Co, NY, New York).
Results
Splenectomy Group
Clinical characteristics of 38 patients who underwent Sx are shown in Table 1. Liver cirrhosis was found in 16 (42.1%) patients in Sx group, and PVT developed in 18 (47.4%). As shown in Table 2, the rate of viral infection, body mass index, and the incidence of Child-Pugh score B were significantly higher in LC-Sx than in non-LC-Sx. In non-LC-Sx, the frequency of DP was significantly higher than that of LC-Sx. The incidence of PVT was significantly higher in LC-Sx than in non-LC-Sx: 68.8% (11 of 16) versus 31.8% (7 of 22; P = .024).
Table 1.
Background of Splenectomy Group.
| Demographics | Total (n = 38) |
|---|---|
| Preoperative factors | |
| Age | 63.5 (45-83) |
| Gender (male/female) | 31/7 |
| LC | 16 (42.1%) |
| HTN | 15 (39.5%) |
| DM | 12 (31.6%) |
| BMI (%) | 22.1 (14.9-30.3) |
| Viral infection (hepatitis B or C) | 14 (36.8%) |
| Child-Pugh score (A/B/C) | 30/8/0 |
| Operative factors | |
| Operation time (min) | 200 (96-650) |
| Blood loss (mL) | 550 (10-8897) |
| Lap/Open | 19 /19 |
| Operation procedures (Sx/DP/Sx with Hx) | 22/13/3 |
| Postoperative factors | |
| Hospital stay (day) | 22 (9-213) |
Abbreviations: BMI, body mass index; DM, diabetes mellitus; HTN, hypertension; LC, liver cirrhosis.
Table 2.
Comparison Between Non-LC-Sx and LC-Sx.
| Groups | Non-LC-Sx (n = 22) | LC--Sx (n = 16) | P Value |
|---|---|---|---|
| Preoperative factors | |||
| Age | 66 (45-83) | 61.5 (51-80) | .258 |
| Gender (male/female) | 17/5 | 14/2 | .422 |
| HTN | 10 | 5 | .376 |
| DM | 9 | 3 | .147 |
| Viral infection (hepatitis B or C) | 4 | 10 | <.01 |
| BMI (%) | 19.6 (14.9-25.4) | 23.9 (16.8-30.3) | <.01 |
| Child-Pugh (A/B/C) | 22/0/0 | 8/8/0 | <.01 |
| Operative factors | |||
| Operation time (min) | 234 (96-650) | 186 (106-548) | .492 |
| Blood loss (mL) | 353 (10-1700) | 1615 (10-8897) | .100 |
| Lap/Open | 10/12 | 9/7 | .511 |
| Operation procedures (Sx/DP/Sx with Hx) | 10/12/0 | 12/1/3 | <.01 |
| Postoperative factors | |||
| Hospital stay (day) | 22.5 (9-213) | 20.5 (10-69) | .510 |
| Intra-abdominal bleeding | 3 | 6 | .088 |
| PVT | 7 (31.8%) | 11 (68.8%) | .024 |
Abbreviations: BMI, body mass index; DM, diabetes mellitus; HTN, hypertension; PVT, portal vein thrombosis.
As shown in Figure 2, platelet counts (×103μL) on preoperative day, POD 1, 3, 5, 7, and 10 were significantly lower in LC-Sx than in non-LC-Sx: 54 (29-260) versus 162 (20-528;P < .01); 96 (18-165) versus 155 (65-442;P < .01); 117 (29-154) versus 157 (77-444;P < .01); 178 (49-258) versus 224 (86-485;P = .033); 242 (21-320) versus 294 (85-644;P = .025); 233 (93-529) versus 467 (10-901,P < .01; Figure 2A). Soluble-form GPVI levels did not show any significant differences between LC-Sx and non-LC-Sx (Figure 2B). Soluble-form GPVI/platelet ratios on preoperative day and POD 1 were significantly higher in LC-Sx than in non-LC-Sx group: 3.31 (0.00-7.38) versus 1.26 (0.00-14.75;P < .01); 2.30 (1.48-15.22) versus 1.64 (0.70-3.86,P < .01; Figure 2C).
Figure 2.
Changes of platelet counts (a), soluble-form GPVI (sGPVI) levels (b), and sGPVI/platelet count ratio (c) in patients with non-LC-Sx and LC-Sx. Open bar, patients with non-LC-Sx; closed bar, patients with LC-Sx. *Versus LC-Sx: P < 0.05, **Versus LC-Sx: P < 0.01.
As shown in Figure 3, VWF levels (U/dL) on preoperative day and POD 5 were significantly higher in LC-Sx than in non-LC-Sx: 313 (0-870) versus 172 (0-776;P = .021); 404 (286-592) versus 280 (165-636,P < .01; Figure 3A). von Willebrand factor propeptide levels (U/dL) on preoperative day and POD 1, 3, 5, 7, 10, and 14 were significantly higher in LC-Sx than in non-LC-Sx: 255 (0-1035) versus 152 (0-551;P < .01); 402 (188-1030) versus 280 (20-657;P < .01); 288 (91-861) versus 193 (9-639;P < .01); 334 (100-582) versus 146 (15-349;P < .01); 255 (102-577) versus 179 (15-357;P = .012); 274 (106-711) versus 172 (20-468;P = .041); 366 (106-939) versus 169 (12-312;P = .019; Figure 3B). A disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 levels did not show any significant differences between LC-Sx and non-LC-Sx (Figure 3C).
Figure 3.
Changes in von Willebrand factor (VWF) levels (a), von Willebrand factor propeptide (VWFpp) levels (b), a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 (ADAMTS13) levels (c) in patients with non-LC-Sx and LC-Sx. Open bar: patients with non-LC-Sx; closed bar, patients with LC-Sx. *Versus LC-Sx: P < 0.05, **Versus LC-Sx: P < 0.01.
Hepatectomy Group
Clinical characteristics of 46 patients who underwent Hx are shown in Table 3. Liver cirrhosis was found in 6 (13.0%) patients in the Hx group, and PVT developed in 6 (13.0%). As shown in Table 4, the rate of viral infection and the incidence of Child-Pugh score B were significantly higher in LC-Hx than in non-LC-Hx. In non-LC-Hx, the incidence of RHx was significantly higher than that of LC-Hx. The incidence of PVT was significantly higher in LC-Hx than in non-LC-Hx: 50.0% (3 of 6) versus 7.5% (3 of 40; P < .01).
Table 3.
Background of Hepatectomy Group.
| Demographics | Total (n = 46) |
|---|---|
| Preoperative factors | |
| Age | 65.5 (23-84) |
| Gender (male/female) | 31/15 |
| LC | 6 (13.0%) |
| HTN | 19 (41.3%) |
| DM | 8 (17.4%) |
| BMI (%) | 20.7 (15.4-34.4) |
| Viral infection (hepatitis B or C) | 7 (15.2%) |
| Child-Pugh score (A/B/C) | 43/3/0 |
| Operative factors | |
| Operation time (min) | 483 (265-919) |
| Blood loss (mL) | 1459 (200-23 340) |
| Lap/Open | 6/40 |
| Operation procedures (Partial/RHx/LHx/L-tri) | 2/27/15/2 |
| Postoperative factors | 32 (12-140) |
| Hospital stay (day) | |
| Intra-abdominal bleeding | 1 (2.2%) |
| PVT | 6 (13.0%) |
Abbreviations: BMI, body mass index; DM, diabetes mellitus; HTN, hypertension; LC, liver cirrhosis; PVT, portal vein thrombosis.
Table 4.
Comparison Between Non-LC-Hx and LC-Hx.
| Groups | Non-LC-Hx (n = 40) | LC--Hx (n = 6) | P Value |
|---|---|---|---|
| Preoperative factors | |||
| Age | 65 (23-84) | 71.5 (64-81) | .064 |
| Gender (male/female) | 26/14 | 5/1 | .372 |
| HTN | 15 | 4 | .176 |
| DM | 6 | 2 | .269 |
| Viral infection (hepatitis B and C) | 1 | 6 | <.01 |
| BMI (%) | 21.1 (15.4-34.4) | 19.9 (16.8-22.8) | .192 |
| Child-Pugh (A/B/C) | 39/1/0 | 4/2/0 | <.01 |
| Operative factors | |||
| Operation time (min) | 477 (285-919) | 486 (265-594) | .534 |
| Blood loss (mL) | 1459 (200-23 340) | 1450 (609-8897) | .660 |
| Lap/Open | 5/35 | 1/5 | .777 |
| Operation procedures (Partial/RHx/LHx/ L-tri) | 0/25/13/2 | 2/2/2/0 | <.01 |
| Postoperative factors | |||
| Hospital stay (day) | 31.5 (12-140) | 33 (22-41) | .812 |
| Bleeding complication | 1 | 0 | .695 |
| PVT | 3 (7.5%) | 3 (50.0%) | <.01 |
Abbreviations: BMI, body mass index, DM, diabetes mellitus; HTN, hypertension; LHx, left hepatectomy; LC, liver cirrhosis; L-tri, left trisectionectomy; PVT, portal vein thrombosis; RHx, right hepatectomy.
As shown in Figure 4, platelet counts (×103μL) on preoperative day were significantly lower in LC-Hx than in non-LC-Hx: 78 (53-148) versus 227 (118-346;P < .01), but platelet counts on postoperative days showed no significant difference between the 2 groups (Figure 4A). Soluble-form GPVI levels (ng/mL) on POD 3 and 5 were significantly higher in LC-Hx than in non-LC-Hx group: 30.8 (13.0-41.2) versus 17.6 (10.7-29.6;P = .014); 26.6 (18.2-32.3) versus 20.5 (10.9-32.8;P = .017; Figure 4B). Glycoprotein VI /platelet ratios on preoperative day and POD 1, 3, 5, 7, and 14 were significantly higher in LC-Hx than in non-LC-Hx group: 2.55 (1.26-7.08) versus 0.87 (0.00-2.96;P < .01); 2.29 (1.77-4.37) versus 1.53 (0.73-4.14;P < .01); 2.72 (1.92-10.30) versus 1.35 (0.60-3.92;P < .01); 2.03 (1.55-7.45) versus 1.47 (0.71-4.83;P = .012); 1.80 (1.41-7.77) versus 1.14 (0.49-3.59;P < .01); 1.41 (1.20-4.93) versus 0.81 (0.61-1.79;P = .046; Figure 4C).
Figure 4.
Changes in platelet counts (a), soluble-form GPVI (sGPVI) levels (b), and sGPVI/platelet count ratio (c) in patients with non-LC-Hx and LC-Hx. Open bar, patients with non-LC-Hx; closed bar: patients with LC-Hx. *Versus LC-Hx: P< 0.05, **Versus LC-Hx: P < 0.01.
As shown in Figure 5, VWF levels (U/dL) did not show any significant differences between LC-Hx and non-LC-Hx (Figure 5A). von Willebrand factor propeptide levels (U/dL) on preoperative day and POD 5 and 14 were significantly higher in LC-Hx than in non-LC-Hx: 320 (183-442) versus 150 (0-809;P = .014); 448 (205-582) versus 218 (22-762;P < .01); 327 (237-564) versus 198 (10-844;P = .027; Figure 5b). A disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 levels showed no significant differences between the 2 groups (Figure 5C).
Figure 5.
Changes in von Willebrand factor (VWF) levels (a), von Willebrand factor propeptide (VWFpp) levels (b), a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 (ADAMTS13) levels (c) in patients with non-LC-Hx and LC-Hx. Open bar, patients with non-LC-Hx; closed bar, patients with LC-Hx. *Versus LC-Hx: P< 0.05, **Versus LC-Hx: P < 0.01.
Comparison Between non-PVT Group and PVT Group
In all 81 patients, PVT developed in 22 patients, and we divided them into 2 groups: PVT group (n = 22) and non-PVT group (n = 59).
As shown in Figure 6, GPVI/platelet ratios on preoperative day and POD 1 were significantly higher in PVT group than in non-PVT group: 2.49 (0.00-14.75) versus 1.00 (0.00-7.08;P < .01); 2.11 (0.85-15.22) versus 1.67 (0.70-4.37;P = .030; Figure 6A). von Willebrand factor levels (U/dL) on preoperative day and POD 3 and 14 were significantly higher in PVT group than in non-PVT group: 314 (0-870) versus 194 (0-776;P < .01); 353 (250-909) versus 311 (141-852;P = .048); 398 (183-813) versus 277 (106-673;P = .043; Figure 6B). von Willebrand factor propeptide levels (U/dL) on preoperative day and POD 5 were significantly higher in PVT group than in non-PVT group: 211 (0-1035) versus 150 (0-809;P = .014); 294 (100-582) versus 205 (15-762;P = .031; Figure 6C). A disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 levels on POD 5, 7, and 10 were significantly higher in PVT group than in non-PVT group: 61.3 (42.1-138.7) versus 49.9 (10.0-124.6;P = .014); 62.5 (28.8-126.1) versus 48.4 (7.5-99.4;P = .012); 71.9 (29.7-124.9) versus 48.3 (3.8-157.5;P = .026; Figure 6D).
Figure 6.
Changes of sGPVI/platelet count ratio (a), von Willebrand factor (VWF) levels (b), von Willebrand factor propeptide (VWFpp) levels (c), a disintegrin-like and metalloproteinase with thrombospondin type 1 motifs 13 (ADAMTS13) levels (d) in patients with non-PVT and patients with PVT. Open bar, patients with non-PVT; closed bar, patients with PVT. *Versus patients with non-PVT: P < 0.05, **Versus patients with non-PVT: P < 0.01.
Discussion
Soluble form of GPVI is recently regarded as specific platelet activation marker. Furthermore, since sGPVI is platelet-specific and the level of sGPVI depends on platelet counts, it is necessary to adjust sGPVI level by platelet counts to evaluate the actual platelet activation.23,29 On the other hand, since endothelial cell injury plays an important role in thrombus formation, VWF and VWFpp have recently received much attention. von Willebrand factor and VWFpp are mainly synthesized by vascular endothelial cells and megakaryocytes and mediate the adhesion of platelets to sites of vascular damage by binding to specific platelet membrane glycoproteins Ib in primary thrombus formation.30–33 von Willebrand factor and VWFpp are stored in vascular endothelial cells and sinusoidal endothelial cell, and they are released from injured or stimulated endothelial cells, reflecting the degree of vascular endothelial damage. Therefore, VWF and VWFpp are considered to be a marker of vascular endothelial cell injury.34 Virchow triad, which is now frequently applied to describe venous thrombosis, consists of 3 factors: alterations in blood flow, injuries to vascular endothelium, and hypercoagulability.35 Because we could not measure alterations in portal venous blood flow, we focused on injuries to vascular endothelium which was estimated by plasma levels of VWF and ADAMTS13, and hypercoagulability which was estimated by plasma levels of sGPVI. In general, activation of platelets is important for arterial thrombosis formation. But Aota et al reported that GPVI level was significantly higher in the patients with DVT than in those without DVT, and they suggested that the activation of platelet was important in venous thrombosis formation.22 Therefore, the activation of the platelet assessed by sGPVI is also considered to affect PVT formation.
In present study, the incidences of PVT in both Sx group and Hx group were significantly higher in patients with LC than in patients with non-LC. In Sx group, in spite of the fact that platelet counts were significantly lower in patients with LC than in patients with non-LC, incidence of PVT was significantly higher in patients with LC than in patients with non-LC (68.8% vs 31.8%). Paying attention to activation of the platelet, sGPVI/platelet ratios on preoperative day and POD 1 were significantly higher in patients with LC than in patients with non-LC, and sGPVI/platelet ratios in patients with LC had decreased gradually after POD 1, whereas its ratios in patients with non-LC remained almost unchanged in perioperative period. These results indicated that patients with LC were already in hypercoagulable state before operation. Pre- and postoperative VWF and VWFpp levels were significantly higher in patients with LC than in patients with non-LC, suggesting that endothelial cell injury chronically occurred in patients with LC. Under the preoperative condition that patients with LC were in hypercoagulable state together with the activation of platelet as well as in injured state of endothelial cells, Sx caused the decrease in portal venous blood flow. It was therefore considered that these 3 factors in patients with LC might lead to early formation of PVT after Sx.
In Hx group, the incidence of PVT was significantly higher in patients with LC than in patients with non-LC (50.0% vs 7.5%). Although preoperative platelet counts were significantly lower in patients with LC than in patients with non-LC, they showed no significant differences between the 2 groups after major Hx. As of GPVI/platelet ratios, patients with LC have had significantly higher levels before and after Hx, compared to patients with non-LC. Pre- and postoperative VWFpp levels were significantly higher in patients with LC than in patients with non-LC, suggesting that endothelial cell injury chronically occurred in patients with LC. Under the preoperative condition that patients with LC were in hypercoagulable state together with the activation of platelet as well as in injured state of endothelial cells, major Hx caused the increase in portal venous pressure and ischemic reperfusion injury in the remnant liver. It was therefore considered that hypercoagulable state, endothelial damage, and alterations in portal venous blood flow in patients with LC might lead to early formation of PVT after major Hx.
Some groups who reported the usefulness of GPVI/platelet ratios as a platelet activation marker enrolled some number of patients with LC in their study.23,29 To dissolve the concern that high GPVI/platelet ratios were a result of patients with LC who might have platelet activation itself and did not predict PVT, we performed the comparison between PVT group and non-PVT group. As a result, GPVI/platelet ratios in pre- and early postoperative period were significantly higher in PVT group than in non-PVT group. In addition, VWF levels and VWFpp levels in perioperative period were significantly higher in PVT group than in non-PVT group. These results indicated that the high GPVI/platelet ratio, VWF and VWFpp levels contributed to develop PVT after hepatobiliary-pancreatic surgery irrespective of cirrhosis status.
In patients with LC, the incidence of PVT did not differ between Sx and Hx: 68.8% versus 50.0%. In contrast, its incidence in patients with non-LC was significantly higher after Sx than after Hx: 31.8% versus 7.5% (P = .013). After Sx in patients with non-LC, platelet counts were significantly and steadily increased until POD 14, while after Hx in patients with non-LC, platelet counts remained unchanged without significant increase, showing significant difference between the 2 groups at POD 3 to 14. Glycoprotein VI /platelet ratios after operation did not show any significant difference between the 2 groups. Taken these results together, it was considered that significant increase in platelet count after Sx might contribute to higher incidence of PVT in non-LC Sx.
In conclusion, patients with LC stay in hypercoagulable state together with activation of platelet and in injured state of endothelial cells before operation, and under these circumstances Sx or Hx causes alterations in portal venous blood flow, resulting in significantly higher incidence of PVT.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
References
- 1. Oomen MW, Bakx R, van Minden M, van Rijn RR, Peters M, Heij HA. Implementation of laparoscopic splenectomy in children and the incidence of portal vein thrombosis diagnosed by ultrasonography. J Pediatr Surg. 2013;48(11):2276–2280. [DOI] [PubMed] [Google Scholar]
- 2. Li M-X, Zhang X-F, Liu Z-W, et al. Risk factors and clinical characteristics of portal vein thrombosis after splenectomy in patients with liver cirrhosis. Hepatobiliary Pancreat Dis Int. 2013;12(5):512–519. [DOI] [PubMed] [Google Scholar]
- 3. Kuboki S, Shimizu H, Ohtsuka M, et al. Incidence, risk factors, and management options for portal vein thrombosis after hepatectomy: a 14-year,single-center experience. Am J Surg. 2015;210(5):878–885. [DOI] [PubMed] [Google Scholar]
- 4. Yoshiya S, Shirabe K, Nakagawara H, et al. Portal vein thrombosis after hepatectomy. World J Surg. 2014;38(6):1491–1497. [DOI] [PubMed] [Google Scholar]
- 5. Ikeda M, Sekimoto M, Takiguchi S, et al. Total splenic vein thrombosis after laparoscopic splenectomy: a possible candidate for treatment. Am J Surg. 2007;193(1):21–25. [DOI] [PubMed] [Google Scholar]
- 6. Ghabril M, Agarwal S, Lacerda M, et al. Portal vein thrombosis is a risk factor for poor early outcomes after liver transplantation: analysis of risk factors and outcomes for portal vein thrombosis in Waitlisted Patients. Transplantation. 2016;100(1):126–133. [DOI] [PubMed] [Google Scholar]
- 7. Kawanaka H, Akahoshi T, Kinjo N, et al. Impact of antithrombin III concentrates on portal vein thrombosis after splenectomy in patients with liver cirrhosis and hypersplenism. Ann Surg. 2010;251(1):76–83. [DOI] [PubMed] [Google Scholar]
- 8. Zocco MA, Di Stasio E, De Cristofaro R, et al. Thrombotic risk factors in patients with liver cirrhosis: correlation with MELD scoring system and portal vein thrombosis development. J Hepatol. 2009;51(4):682–689. [DOI] [PubMed] [Google Scholar]
- 9. Tripodi A, Primignani M, Lemma L, et al. Evidence that low protein C contributes to the procoagulant imbalance in cirrhosis. J Hepatol 2013;59(2):265–270. [DOI] [PubMed] [Google Scholar]
- 10. Valla DC. Thrombosis and anticoagulation in liver disease. Hepatology. 2008;47(4):1384–1393. [DOI] [PubMed] [Google Scholar]
- 11. Delahousse B, Labat-Debelleix V, Decalonne L, et al. Comparative study of coagulation and thrombin generation in the portal and jugular plasma of patients with cirrhosis. Thromb Haemost. 2010;104(4):741–749. [DOI] [PubMed] [Google Scholar]
- 12. Moroi A, Watson S. Impact of the PI3-kinase/Akt pathway on ITAM and hemITAM receptors: haemostasis, platelet activation and antithrombotic therapy. Biochemical pharmacology. 2015;94(3):186–194. [DOI] [PubMed] [Google Scholar]
- 13. Tsuji M, Ezumi Y, Arai M, Takayama H. A novel association of Fc receptor gamma-chain with glycoprotein VI and their coexpression as a collagen receptor in human platelets. J Biol Chem. 1997;272(38):23528–23531. [DOI] [PubMed] [Google Scholar]
- 14. Leitinger B. Transmembrane collagen receptors. Annu Rev Cell Dev Biol. 2011;27:265–290. [DOI] [PubMed] [Google Scholar]
- 15. Gardiner EE, Arthur JF, Kahn ML, Berndt MC, Andrews RK. Regulation of platelet membrane levels of glycoprotein VI by a platelet-derived metalloproteinase. Blood. 2004;104(12):3611–3617. [DOI] [PubMed] [Google Scholar]
- 16. Gardiner EE, Karunakaran D, Shen Y, Arthur JF, Andrews RK, Berndt MC. Controlled shedding of platelet glycoprotein (GP)VI and GPIb-IX-V by ADAM family metalloproteinases. J Thromb Haemost. 2007;5(7):1530–1537. [DOI] [PubMed] [Google Scholar]
- 17. Gardiner EE, Al-Tamimi M, Andrews RK, Berndt MC. Platelet receptor shedding. Methods Mol Biol. 2012;788:321–339. [DOI] [PubMed] [Google Scholar]
- 18. Al-Tamimi M, Arthur JF, Gardiner E, Andrews RK. Focusing on plasma glycoprotein VI. Thromb Haemost. 2012;107(4):648–655. [DOI] [PubMed] [Google Scholar]
- 19. Bigalke B, Stellos K, Weig HJ, et al. Regulation of platelet glycoprotein VI (GPVI) surface expression and of soluble GPVI in patients with atrial fibrillation (AF) and acute coronary syndrome (ACS). Basic Res Cardiol. 2009;104(3):352–357. [DOI] [PubMed] [Google Scholar]
- 20. Al-Tamimi M, Gardiner EE, Thom JY, et al. Soluble glycoprotein VI is raised in the plasma of patients with acute ischemic stroke. Stroke. 2011;42(2):498–500. [DOI] [PubMed] [Google Scholar]
- 21. Al-Tamimi M, Grigoriadis G, Tran H, et al. Coagulation-induced shedding of platelet glycoprotein VI mediated by factor Xa. Blood. 2011;117(14):3912–3920. [DOI] [PubMed] [Google Scholar]
- 22. Aota T, Naitoh K, Wada H, et al. Elevated soluble platelet glycoprotein VI is a useful marker for DVT in postoperative patients treated with edoxaban. Int J Hematol. 2014;100(5):450–456. [DOI] [PubMed] [Google Scholar]
- 23. Egan K, Dillon A, Dunne E, et al. Increased soluble GPVI levels in cirrhosis: evidence for early in vivo platelet activation. J Thromb Thrombolysis. 2017;43(1):54–59. [DOI] [PubMed] [Google Scholar]
- 24. Takayama H, Hosaka Y, Nakayama K. A novel antiplatelet antibody therapy that induces cAMP-dependent endocytosis of the GPVI/Fc receptor gamma-chain complex. J Clin Invest. 2008;118(5):1785–1795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Yamashita Y, Naitoh K, Wada H, et al. Elevated plasma levels of soluble platelet glycoprotein VI (GPVI) in patients with thrombotic microangiopathy. Thromb Res. 2014;133(3):440–444. [DOI] [PubMed] [Google Scholar]
- 26. Takahashi N, Wada H, Usui M, et al. Behavior of ADAMTS13 and Von Willebrand factor levels in patients after living donor liver transplantation. Thromb Res. 2013;131(3):225–229. [DOI] [PubMed] [Google Scholar]
- 27. Kokame K, Nobe Y, Kokubo Y, et al. FRETSVWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol. 2005;129(1):93–100. [DOI] [PubMed] [Google Scholar]
- 28. Nangliya V, Sharma A, Yadav D, et al. Study of trace elements in liver cirrhosis patients and their role in prognosis of disease. Biol Trace Elem Res. 2015;165:35–40. [DOI] [PubMed] [Google Scholar]
- 29. Takahashi N, Usui M, Naitoh K, et al. Elevated soluble platelet glycoprotein VI levels in patients after living donor liver transplantation. Clin Appl Thromb Hemost. 2017;23(3):274–281. [DOI] [PubMed] [Google Scholar]
- 30. Soejima K, Mimura N, Hirashima M, et al. A novel human metalloprotease synthesized in the liver and secreted into the blood: possibly, the von Willebrand factor-cleaving protease? J Biochem. 2001;130(4):475–480. [DOI] [PubMed] [Google Scholar]
- 31. Levy GG, Nichols WC, Lian EC, et al. Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura. Nature. 2001;413(6855):488–494. [DOI] [PubMed] [Google Scholar]
- 32. Zheng X, Chung D, Takayama TK, Majerus EM, Sadler JE, Fujikawa K. Structure of von Willebrand factor-cleaving protease (ADAMTS 13), metalloproteinase involved in thrombotic thrombocytopenic purpura. J Biol Chem. 2001;276(44):41059–41063. [DOI] [PubMed] [Google Scholar]
- 33. Plaimauer B, Zimmermann K, Völkel D, et al. Cloning, expression and characterization of the von Willebrand factor-cleaving protease (ADAMTS 13). Blood. 2002;100(10):3626–3632. [DOI] [PubMed] [Google Scholar]
- 34. Ito-Habe N, Wada H, Matsumoto T, et al. Elevated Von Willebrand factor propeptide for the diagnosis of thrombotic microangiopathy and for predicting a poor outcome. Int J Hematol. 2011;93(1):47–52. [DOI] [PubMed] [Google Scholar]
- 35. Dickson BC. Venous thrombosis: on the history of Virchow’s triad. Univ Toronto Med J. 2004;81:166–171. [Google Scholar]