Association between the severity of acquired von Willebrand syndrome and gastrointestinal bleeding after continuous-flow left ventricular assist device implantation

Association between the severity of acquired von Willebrand syndrome and gastrointestinal... Abstract OBJECTIVES Acquired von Willebrand syndrome, characterized by the reduction in von Willebrand factor (vWF) large multimers, has recently been considered as one of the causes of gastrointestinal bleeding (GIB). It remains unclear whether its haematological severity is linked with susceptibility to bleeding because the definition of the haematological severity of acquired von Willebrand syndrome has not been precisely determined. This study sought to establish a quantitative methodology to assess the haematological severity of acquired von Willebrand syndrome and to define the threshold for occurrence of GIB in patients implanted with left ventricular assist devices (LVADs). METHODS In total, 41 patients treated with continuous-flow LVAD implanted between 2011 and 2017 at Tohoku University Hospital were investigated. vWF large multimers were quantitatively evaluated using the ‘vWF large multimer index’ defined as the ratio of a large multimer proportion in total vWF derived from a patient to that from a normal control. Using this index, the amount of vWF large multimers was expressed as a percentage of its normal control value obtained with a simultaneous analysis of each time measurement. RESULTS Twelve (29%) patients developed GIB events during follow-up periods (median 591 days) after an LVAD implantation. The vWF large multimer index in patients with GIB was significantly lower than that in those without GIB (25.0 ± 10.3% vs 37.5 ± 17.8%, P = 0.008). Most importantly, all patients experiencing GIB exhibited a vWF large multimer index below 40%. CONCLUSIONS Patients with GIB exhibited a more severe loss of vWF large multimers. The vWF large multimer index may dictate the risk of GIB after an LVAD implantation. Clinical trial registration number UMIN000018135. Acquired von Willebrand syndrome, Left ventricular assist device, Gastrointestinal bleeding INTRODUCTION Left ventricular assist devices (LVADs) contribute to the treatment of end-stage heart failure [1, 2]. Bleeding complications, in particular gastrointestinal bleeding (GIB) [3], occur in 10–33% of patients with continuous-flow types of LVADs [4, 5]. Recently, acquired von Willebrand syndrome (aVWS) induced by LVADs has been highlighted as a possible cause of GIB [6, 7]. von Willebrand factor (vWF) plays a critical role in haemostasis. It is cleaved into multimers composed of 2–80 subunits by ADAMTS13 (a disintegrin and metalloproteinase with thrombospondin Type 1 motifs, member 13), an enzyme with specific vWF-cleaving activity in a shear stress-dependent manner [8]. High molecular weight multimers play important roles in haemostasis [9]. Unphysiologically high shear stress inside pumps induces excessive cleavage of vWF multimers, leading to a loss of such large multimers and a haemostatic disorder known as aVWS [10, 11]. The degree of loss of vWF large multimers has been variably reported, since a quantitative methodology to define the loss of vWF large multimers has not been fully established. Therefore, it remains undetermined as to whether the severity of haematological aVWS is linked with LVAD-associated GIB [12]. To address this question, we quantitatively analysed vWF large multimers in 41 patients after an LVAD implantation, introducing the vWF large multimer index [13] and evaluated the association between the haematological severity of aVWS and the occurrence of GIB complications. METHODS Patients evaluated in this study From July 2011 to June 2017, 47 patients underwent implantation of LVADs in Tohoku University Hospital. Six patients were excluded from the study, because they had undergone heart transplantation or passed away before the installation of vWF multimer analysis. Thus, 41 patients participated in the study. We performed vWF multimer analysis and followed patient clinical courses until June 2017. All 41 patients were treated with anticoagulant therapy with warfarin targeting prothrombin time–international normalized ratio (PT-INR) levels between 2.0 and 3.0, and also by antiplatelet therapy, principally with aspirin. Since 8 patients were aspirin intolerant, 7 of these were treated with an adenosine diphosphate receptor blocker clopidogrel, whereas 1 was treated with a phosphodiesterase 3-inhibitor cilostazol. At the early postoperative period, we titrated the dose of heparin targeting activated partial thromboplastin time (APTT) levels between 45 and 55 s until warfarin control was established. vWF large multimers were evaluated in our laboratory on Day 200 (median; range Day 3–1254) after an LVAD implantation under stable conditions rather than in the acute phase of bleeding. ADAMTS13 activity was measured at the same time, based on the FRETS-VWF73 assay [14] by SRL Co. (Tokyo, Japan). For the analysis, plasma was isolated in a platelet-free manner within 1 h after blood extraction and stored at −80°C until analysis. Quantification of large multimers of von Willebrand factor We first determined the concentrations of vWF in the plasma of the normal control subject and each patient by Western blot analysis under reduced conditions followed by densitometric analysis. The control plasma was prepared by extracting plasma from a single healthy person. A dispensed sample from the control plasma was consistently used for each time measurement throughout the entire study. The vWF antigen levels in the plasma of patient and control were determined by Western blot in a reduced condition. Then, plasma containing the same amount of vWF in each patient as that of the control was analysed in adjacent lanes by the vWF multimer analysis performed essentially as described [15] with polyclonal rabbit anti-human vWF antibody (DAKO, Glostrup, Denmark) as the primary antibody. The Western blots probing vWF were finally visualized by ImmunoStar® Zeta chemiluminescence (Wako, Osaka, Japan) and evaluated by Image Quant LAS 4000 mini (GE Healthcare) and ImageJ as described previously [13]. Bands 1–5 (starting from the lowest molecular weight band) were classified as low, 6–10 as medium and all those >10 as large multimers. The vWF large multimer proportion was calculated as the densitometric area of those bands. The vWF large multimer index was defined as the ratio of the large multimer proportion in total vWF derived from a patient to that from a normal control analysed in the adjacent lane [13]. Using this index, the quantity of vWF large multimers in a patient was expressed as a percentage of its normal control value obtained with simultaneous analysis of each time measurement (Fig. 1). Figure 1: View largeDownload slide Quantification method for calculation of the vWF large multimer index. vWF large multimer index in a patient is expressed as a percentage of the proportion of vWF large multimer in a normal control. LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 1: View largeDownload slide Quantification method for calculation of the vWF large multimer index. vWF large multimer index in a patient is expressed as a percentage of the proportion of vWF large multimer in a normal control. LVAD: left ventricular assist device; vWF: von Willebrand factor. Measurement of plasma von Willebrand factor antigen level and ristocetin cofactor activity vWF antigen (vWF: Ag) was measured by SRL Co. using the latex agglutination reaction with STA®-LIATEST VWF: Ag (Roche Diagnostics, Switzerland). vWF ristocetin cofactor activity (vWF: RCo) was also analysed by SRL Co. with BC von Willebrand Reagent (Siemens, Germany). Gastrointestinal bleeding events GIB was defined as any clinically suspected or documented suspicion of bleeding from the gastrointestinal tract as indicated with a de novo fall in haemoglobin level and the appearance of melena, haematochezia, haematemesis or guaiac positive stools as described previously [4, 16]. Statistical analyses All statistical analyses were performed using JMP® Pro version 13 (SAS Institute Inc., Cary, NC, USA). Continuous variables are expressed as mean ± standard deviation. Categorical variables are presented as frequency and percentages, and inter-group comparisons of categorical variables were analysed using Fisher’s exact test or χ2 test. If a value in a cell was <5, the Fisher’s exact test was applied. Intergroup comparison of continuous variables was performed using the Student’s t-test or the Welch’s t-test, if normally distributed; otherwise, Mann–Whitney U-test was used. Shapiro–Wilk test was applied to determine whether the variable follows a normal distribution. Kaplan–Meier methods were used to estimate GIB event-free survival. The median follow-up time was computed using the inverse Kaplan–Meier method. Receiver operating characteristic curves (ROC) were performed to identify the cut-off value of a vWF large multimer index. The difference was considered significant at a P-value <0.05. Ethics This study was performed in accordance with the Declaration of Helsinki. The study protocol was approved by the Institutional Ethics Committee in Tohoku University (Certification number; 2017-1-386). Written informed consent was obtained from each patient. RESULTS Patient characteristics Patient demographics and clinical information categorized by the 2 different pump types are summarized in Table 1. All patients were followed for a median of 591 days, and 28 patients were supported with LVADs throughout the follow-up periods. The remaining 13 patients were followed until the occurrence of the following events: heart transplantation in 5 patients on Days 591, 999, 1156, 1296 and 1550 after LVAD implantation; LVAD explantation in 1 patient due to severe device infection on Day 834; pump exchange in 4 patients on Days 384, 681, 778 and 434; and death in 3 patients on Days 248, 892 and 1361, respectively. No significant differences were observed in age, sex or aetiology between patients treated with axial (n = 33) and centrifugal (n = 8) types of LVADs (Table 1). Table 1: Characteristics of patients implanted with axial- and centrifugal-type LVADs LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; LVAD: left ventricular assist device; rpm: revolutions per minute; SD: standard deviation. Table 1: Characteristics of patients implanted with axial- and centrifugal-type LVADs LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; LVAD: left ventricular assist device; rpm: revolutions per minute; SD: standard deviation. Gastrointestinal bleeding complications While thrombotic events of cerebral infarction, superior mesenteric artery thrombosis and pump thrombosis occurred in 7 patients and severe device infection occurred in 3 patients during the follow-up periods, GIB complications occurred in 12 (29%) patients at the median LVAD support time of 179 days (range 14–1106 days) (Tables 1 and 2). Ten GIB patients were supported with axial-type LVADs, and the other 2 with centrifugal-type LVADs. Mean haemoglobin levels in the acute bleeding phase of the 12 patients were 77.9 ± 17.1 g/l, suggesting that relatively large amounts of blood were lost on the bleeding events, whereas their PT-INR and APTT were within therapeutic ranges. The GIB-free survival curve indicated a 6-month event-free survival of 85.3%, 1-year survival of 77.0% and 2-year survival of 73.0% by the Kaplan–Meier method (Fig. 2). Four patients (Cases 1–4, Table 2) developed GIB in the early periods of 14–40 days after LVAD implantation. In 2 of these patients, bleeding lesions were detected by endoscopy. The bleeding sites contained oozing haemorrhages from ulcers and/or erosions, and angiodysplasia was not detected in either patient. On the other hand, GIB events occurred in the remaining 8 patients (Cases 5–12, Table 2) in the later periods of Day 142 and after. Upper and/or lower gastrointestinal examinations did not identify distinct bleeding lesions except for 2 patients. The origin of both bleeds in Case 5 at Day 142 and in Case 9 at Day 266 was confirmed to be angiodysplasia. Table 2: Clinical characteristics and course of 12 patients with GIB   Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected    Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected  a The data were examined in the acute phase of bleeding. b vWF multimer analysis and measurement of activity of ADAMTS13 were performed with patients in the stable state. c Case 3 was presented previously [17]. ADAMTS13: a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; Hb: haemoglobin; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; rpm: revolutions per minute; vWF: von Willebrand factor; vWF: RCo/vWF: Ag: ristocetin cofactor activity: von Willebrand factor antigen ratio. Table 2: Clinical characteristics and course of 12 patients with GIB   Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected    Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected  a The data were examined in the acute phase of bleeding. b vWF multimer analysis and measurement of activity of ADAMTS13 were performed with patients in the stable state. c Case 3 was presented previously [17]. ADAMTS13: a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; Hb: haemoglobin; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; rpm: revolutions per minute; vWF: von Willebrand factor; vWF: RCo/vWF: Ag: ristocetin cofactor activity: von Willebrand factor antigen ratio. Figure 2: View largeDownload slide Actuarial freedom from first gastrointestinal bleeding after continuous-flow type left ventricular assist device implantation. GIB: gastrointestinal bleeding. Figure 2: View largeDownload slide Actuarial freedom from first gastrointestinal bleeding after continuous-flow type left ventricular assist device implantation. GIB: gastrointestinal bleeding. The characteristics of the LVAD patients with or without GIB are summarized in Table 3. No significant differences were observed in gender, age, aetiology of heart disease, pump types or haematological data between the 2 groups using the Fisher’s exact test. All the other laboratory findings were not significantly different, except for creatinine (serum creatinine levels: GIB groups 1.29 ± 0.68 vs Non-GIB groups 0.74 ± 0.31 mg/dl, P = 0.017) by the Student’s t-test or the Welch’s t-test, when appropriate. Table 3: Characteristics of LVAD Patients with or without GIB   Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48    Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48  APTT: activated partial thromboplastin time; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; SD: standard deviation. Table 3: Characteristics of LVAD Patients with or without GIB   Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48    Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48  APTT: activated partial thromboplastin time; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; SD: standard deviation. Evaluation of vWF large multimers by the vWF large multimer index All LVAD patients exhibited severe loss of vWF large multimers [18], with a mean index of 33.8 ± 16.8%. With regard to pump types, vWF large multimer indices in patients with axial-type LVADs were significantly lower compared to those with centrifugal-type LVADs (31.4 ± 16.7% vs 43.8 ± 14.2%, P = 0.030) by the Student’s t-test (Fig. 3). Those in patients with GIB were significantly lower compared to those without GIB (25.0 ± 10.3% vs 37.5 ± 17.8%, P = 0.008) by the Welch’s t-test (Fig. 4). Notably, all the patients with GIB exhibited indices below 40%, and no patients with an index above 40% developed GIB. Furthermore, 12 (41%) of the 29 patients with indices below 40% experienced GIB during the median follow-up period of 591 days. ROC identified the cut-off value for vWF large multimer on GIB as 39.4% (sensitivity 100%, specificity 41.4%) (see Supplementary Material, Fig. S1). On the other hand, the other parameters including vWF: Ag (103.6 ± 50.4% vs 116.4 ± 57.7%, P = 0.74), vWF: RCo (49.8 ± 39.0% vs 61.3 ± 44.2%, P = 0.78), vWF: RCo/vWF: Ag (0.52 ± 0.28 vs 0.42 ± 0.22, P = 0.85) and ADAMTS13 activities (120.6 ± 35.3% vs 125.5 ± 36.6%, P = 0.69) were not statistically different (GIB group vs Non-GIB group, respectively) by the Student’s t-test or the Welch’s t-test, when appropriate. Figure 3: View largeDownload slide Comparison of vWF large multimer indices in patients with axial-type LVADs (n = 33) and centrifugal-type LVADs (n = 8). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. Cases who developed gastrointestinal bleeding are circled in red. GIB: gastrointestinal bleeding; LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 3: View largeDownload slide Comparison of vWF large multimer indices in patients with axial-type LVADs (n = 33) and centrifugal-type LVADs (n = 8). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. Cases who developed gastrointestinal bleeding are circled in red. GIB: gastrointestinal bleeding; LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 4: View largeDownload slide Comparison of vWF large multimer indices in left ventricular assist device patients with (n = 12) and without GIB (n = 29). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. GIB: gastrointestinal bleeding; vWF: von Willebrand factor. Figure 4: View largeDownload slide Comparison of vWF large multimer indices in left ventricular assist device patients with (n = 12) and without GIB (n = 29). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. GIB: gastrointestinal bleeding; vWF: von Willebrand factor. DISCUSSION In this study, we have quantitatively evaluated the loss of vWF large multimers that could indicate the haematological severity of aVWS in 41 patients after an LVAD implantation. We have demonstrated that GIB occurred in 12 (29%) patients who exhibited a more severe loss of vWF large multimers than those without GIB. Further, cut-off value for vWF large multimer index on GIB was 39.4% by ROC analysis. Thus, aVWS in a severe form associated with a certain threshold of multimer index may dictate the risk of LVAD-associated GIB during the course of follow-up. In some previous studies, vWF large multimers were evaluated by vWF large multimer ratios [13, 19, 20]. In a study by Tamura et al. [13] analysing 31 severe aortic stenosis patients, vWF large multimer ratios were decreased in a pressure gradient-dependent manner. Their study demonstrated that the ratios were distributed predominantly between 10% and 35%, although these ratios overlapped extensively with those of normal controls analysed simultaneously [13]. In another study, vWF large multimer ratios have also been shown to decrease depending on the severity of aortic stenosis where by ratios were distributed between 4% and 14% [19]. Furthermore, Meyer et al. [20] demonstrated that in 102 patients treated with LVAD, pump speed was weakly, but significantly, correlated inversely with vWF large multimer ratios distributed between 20% and 60%. It has been reported that vWF multimer profiles do not correlate with the incidence of bleeding complications [20]. Thus, although these ratios may indicate some association with clinical parameters, the absolute values of the vWF large multimer ratios vary considerably between studies. Therefore, it would be difficult to use the ratios for comparison across studies. ROC vWF large multimer analysis is categorized as a manoeuvre of Western blotting [15], the results obtained would appear to be influenced by the vWF antigen amount to be analysed and the exposure time in the final visualization procedure. Variability of vWF large multimer ratios may be attributable to those factors associated with laboratory work. To avoid such variability, we have recently proposed a novel value, namely the vWF large multimer index [13], since the index is a relative value using vWF large multimers of a normal control as a reference. We further confirmed using our quantitative method that the axial pump group exhibited more severe loss of vWF large multimers than the centrifugal pump group as shown previously [20, 21]. It is notable that 39.4% could be a predictive value for subsequent GIB in patients treated with continuous-flow-type LVADs, irrespective of either axial- or centrifugal-type pump involvement. This index may also be useful in avoiding GIB complications by titrating the intensity of antithrombotic therapy, since vWF is a platelet-binding factor and plays a critical role in haemostasis. It may be worthwhile to consider modifying an the antithrombotic regimen for patients developing severe aVWS, who exhibit a vWF large multimer index lower than 40%. However, since the number of patients in this study was small, a modification of the anticoagulation and antiplatelet therapy protocol in patients with LVAD-associated aVWS warrants further larger scale clinical investigation to establish conclusive evidence. The causes of GIB are clinically considered to be multifactorial. Having severe aVWS as a substrate, development of gastrointestinal ulcer at an early stage and angiodysplasia at a later stage may evoke GIB as a superimposed factor in patients treated with LVAD. In our present study, the majority of GIB patients did not have a clear cause of GIB, and only 4 patients were clearly found to have bleeding lesions such as ulcer, erosion and angiodysplasia. Among those 4 patients the mean value of the vWF multimer index was 25.2%, whereas the patients without clear causes of GIB exhibited a mean index of 28.8%. There were no significant differences (P = 0.67) in those values. The correlation between the vWF large multimer index and development of gastrointestinal organic lesion was not prove, so further study will be required to clarify this issue. Limitations The limitations to this study include the fact that it was a retrospective single-centre study with a small number of patients. Further study is required for unequivocal establishment of the vWF large multimer index as a predictive measure for GIB complications in patients implanted with LVADs. CONCLUSIONS The vWF large multimer index, which is a novel quantitative methodology employed to evaluate the haematological severity of aVWS, was applied in all 41 patients after LVAD implantation. Using this index, patients with GIB exhibited more severe aVWS compared to those without GIB haematologically. GIB patients exhibited severe aVWS with large multimer indices of approximately ≤40%. The vWF large multimer index may predict a risk of GIB after an LVAD implantation. SUPPLEMENTARY MATERIAL Supplementary material is available at EJCTS online. FUNDING This work was supported by JSPS KAKENHI [15K10204 to S.K., M.A. and Y.S.] and partly supported by a Health and Labor Sciences Research Grant for Research on rare and intractable diseases from the Ministry of Health, Labor and Welfare, Japan, to H.H. and Y.S., and grants from the SENSHIN Medical Research Foundation and the Suzuken Memorial Foundation to H.H. Conflict of interest: none declared. REFERENCES 1 Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W et al.   Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med  2001; 345: 1435– 43. Google Scholar CrossRef Search ADS PubMed  2 Hata H, Fujita T, Shimahara Y, Sato S, Yanase M, Seguchi O et al.   Early and mid-term outcomes of left ventricular assist device implantation and future prospects. Gen Thorac Cardiovasc Surg  2015; 63: 557– 64. Google Scholar CrossRef Search ADS PubMed  3 Suarez J, Patel CB, Felker GM, Becker R, Hernandez AF, Rogers JG. Mechanisms of bleeding and approach to patients with axial-flow left ventricular assist devices. Circ Heart Fail  2011; 4: 779– 84. Google Scholar CrossRef Search ADS PubMed  4 Draper KV, Huang RJ, Gerson LB. GI bleeding in patients with continuous-flow left ventricular assist devices: a systematic review and meta-analysis. Gastrointest Endosc  2014; 80: 435– 46.e1. Google Scholar CrossRef Search ADS PubMed  5 Demirozu ZT, Radovancevic R, Hochman LF, Gregoric ID, Letsou GV, Kar B et al.   Arteriovenous malformation and gastrointestinal bleeding in patients with the HeartMate II left ventricular assist device. J Heart Lung Transplant  2011; 30: 849– 53. Google Scholar CrossRef Search ADS PubMed  6 Uriel N, Pak SW, Jorde UP, Jude B, Susen S, Vincentelli A et al.   Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation. J Am Coll Cardiol  2010; 56: 1207– 13. Google Scholar CrossRef Search ADS PubMed  7 Geisen U, Heilmann C, Beyersdorf F, Benk C, Berchtold-Herz M, Schlensak C et al.   Non-surgical bleeding in patients with ventricular assist devices could be explained by acquired von Willebrand disease. Eur J Cardiothorac Surg  2008; 33: 679– 84. Google Scholar CrossRef Search ADS PubMed  8 Shida Y, Nishio K, Sugimoto M, Mizuno T, Hamada M, Kato S et al.   Functional imaging of shear-dependent activity of ADAMTS13 in regulating mural thrombus growth under whole blood flow conditions. Blood  2008; 111: 1295– 8. Google Scholar CrossRef Search ADS PubMed  9 Matsumoto M, Kawaguchi S, Ishizashi H, Yagi H, Iida J, Sakaki T et al.   Platelets treated with ticlopidine are less reactive to unusually large von Willebrand factor multimers than are those treated with aspirin under high shear stress. Pathophysiol Haemost Thromb  2005; 34: 35– 40. Google Scholar CrossRef Search ADS PubMed  10 Malehsa D, Meyer AL, Bara C, Struber M. Acquired von Willebrand syndrome after exchange of the HeartMate XVE to the HeartMate II ventricular assist device. Eur J Cardiothorac Surg  2009; 35: 1091– 3. Google Scholar CrossRef Search ADS PubMed  11 Nascimbene A, Neelamegham S, Frazier OH, Moake JL, Dong JF. Acquired von Willebrand syndrome associated with left ventricular assist device. Blood  2016; 127: 3133– 41. Google Scholar CrossRef Search ADS PubMed  12 Baghai M, Heilmann C, Beyersdorf F, Nakamura L, Geisen U, Olschewski M et al.   Platelet dysfunction and acquired von Willebrand syndrome in patients with left ventricular assist devices. Eur J Cardiothorac Surg  2015; 48: 421– 7. Google Scholar CrossRef Search ADS PubMed  13 Tamura T, Horiuchi H, Imai M, Tada T, Shiomi H, Kuroda M et al.   Unexpectedly high prevalence of acquired von Willebrand syndrome in patients with severe aortic stenosis as evaluated with a novel large multimer index. J Atheroscler Thromb  2015; 22: 1115– 23. Google Scholar CrossRef Search ADS PubMed  14 Kokame K, Nobe Y, Kokubo Y, Okayama A, Miyata T. FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol  2005; 129: 93– 100. Google Scholar CrossRef Search ADS PubMed  15 Ruggeri ZM, Zimmerman TS. The complex multimeric composition of factor VIII/von Willebrand factor. Blood  1981; 57: 1140– 3. Google Scholar PubMed  16 Goldstein DJ, Aaronson KD, Tatooles AJ, Silvestry SC, Jeevanandam V, Gordon R et al.   Gastrointestinal bleeding in recipients of the HeartWare Ventricular Assist System. JACC Heart Fail  2015; 3: 303– 13. Google Scholar CrossRef Search ADS PubMed  17 Sakatsume K, Akiyama M, Saito K, Kawamoto S, Horiuchi H, Saiki Y. Intractable bleeding tendency due to acquired von Willebrand syndrome after Jarvik 2000 implant. J Artif Organs  2016; 19: 289– 92. Google Scholar CrossRef Search ADS PubMed  18 Heilmann C, Geisen U, Beyersdorf F, Nakamura L, Trummer G, Berchtold-Herz M et al.   Acquired Von Willebrand syndrome is an early-onset problem in ventricular assist device patients. Eur J Cardiothorac Surg  2011; 40: 1328– 33; discussion 233. Google Scholar PubMed  19 Vincentelli A, Susen S, Le Tourneau T, Six I, Fabre O, Juthier F et al.   Acquired von Willebrand syndrome in aortic stenosis. N Engl J Med  2003; 349: 343– 9. Google Scholar CrossRef Search ADS PubMed  20 Meyer AL, Malehsa D, Budde U, Bara C, Haverich A, Strueber M. Acquired von Willebrand syndrome in patients with a centrifugal or axial continuous flow left ventricular assist device. JACC Heart Fail  2014; 2: 141– 5. Google Scholar CrossRef Search ADS PubMed  21 Netuka I, Kvasnicka T, Kvasnicka J, Hrachovinova I, Ivak P, Marecek F et al.   Evaluation of von Willebrand factor with a fully magnetically levitated centrifugal continuous-flow left ventricular assist device in advanced heart failure. J Heart Lung Transplant  2016; 35: 860– 7. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Association between the severity of acquired von Willebrand syndrome and gastrointestinal bleeding after continuous-flow left ventricular assist device implantation

Loading next page...
 
/lp/ou_press/association-between-the-severity-of-acquired-von-willebrand-syndrome-EQpOKBCuEi
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezy172
Publisher site
See Article on Publisher Site

Abstract

Abstract OBJECTIVES Acquired von Willebrand syndrome, characterized by the reduction in von Willebrand factor (vWF) large multimers, has recently been considered as one of the causes of gastrointestinal bleeding (GIB). It remains unclear whether its haematological severity is linked with susceptibility to bleeding because the definition of the haematological severity of acquired von Willebrand syndrome has not been precisely determined. This study sought to establish a quantitative methodology to assess the haematological severity of acquired von Willebrand syndrome and to define the threshold for occurrence of GIB in patients implanted with left ventricular assist devices (LVADs). METHODS In total, 41 patients treated with continuous-flow LVAD implanted between 2011 and 2017 at Tohoku University Hospital were investigated. vWF large multimers were quantitatively evaluated using the ‘vWF large multimer index’ defined as the ratio of a large multimer proportion in total vWF derived from a patient to that from a normal control. Using this index, the amount of vWF large multimers was expressed as a percentage of its normal control value obtained with a simultaneous analysis of each time measurement. RESULTS Twelve (29%) patients developed GIB events during follow-up periods (median 591 days) after an LVAD implantation. The vWF large multimer index in patients with GIB was significantly lower than that in those without GIB (25.0 ± 10.3% vs 37.5 ± 17.8%, P = 0.008). Most importantly, all patients experiencing GIB exhibited a vWF large multimer index below 40%. CONCLUSIONS Patients with GIB exhibited a more severe loss of vWF large multimers. The vWF large multimer index may dictate the risk of GIB after an LVAD implantation. Clinical trial registration number UMIN000018135. Acquired von Willebrand syndrome, Left ventricular assist device, Gastrointestinal bleeding INTRODUCTION Left ventricular assist devices (LVADs) contribute to the treatment of end-stage heart failure [1, 2]. Bleeding complications, in particular gastrointestinal bleeding (GIB) [3], occur in 10–33% of patients with continuous-flow types of LVADs [4, 5]. Recently, acquired von Willebrand syndrome (aVWS) induced by LVADs has been highlighted as a possible cause of GIB [6, 7]. von Willebrand factor (vWF) plays a critical role in haemostasis. It is cleaved into multimers composed of 2–80 subunits by ADAMTS13 (a disintegrin and metalloproteinase with thrombospondin Type 1 motifs, member 13), an enzyme with specific vWF-cleaving activity in a shear stress-dependent manner [8]. High molecular weight multimers play important roles in haemostasis [9]. Unphysiologically high shear stress inside pumps induces excessive cleavage of vWF multimers, leading to a loss of such large multimers and a haemostatic disorder known as aVWS [10, 11]. The degree of loss of vWF large multimers has been variably reported, since a quantitative methodology to define the loss of vWF large multimers has not been fully established. Therefore, it remains undetermined as to whether the severity of haematological aVWS is linked with LVAD-associated GIB [12]. To address this question, we quantitatively analysed vWF large multimers in 41 patients after an LVAD implantation, introducing the vWF large multimer index [13] and evaluated the association between the haematological severity of aVWS and the occurrence of GIB complications. METHODS Patients evaluated in this study From July 2011 to June 2017, 47 patients underwent implantation of LVADs in Tohoku University Hospital. Six patients were excluded from the study, because they had undergone heart transplantation or passed away before the installation of vWF multimer analysis. Thus, 41 patients participated in the study. We performed vWF multimer analysis and followed patient clinical courses until June 2017. All 41 patients were treated with anticoagulant therapy with warfarin targeting prothrombin time–international normalized ratio (PT-INR) levels between 2.0 and 3.0, and also by antiplatelet therapy, principally with aspirin. Since 8 patients were aspirin intolerant, 7 of these were treated with an adenosine diphosphate receptor blocker clopidogrel, whereas 1 was treated with a phosphodiesterase 3-inhibitor cilostazol. At the early postoperative period, we titrated the dose of heparin targeting activated partial thromboplastin time (APTT) levels between 45 and 55 s until warfarin control was established. vWF large multimers were evaluated in our laboratory on Day 200 (median; range Day 3–1254) after an LVAD implantation under stable conditions rather than in the acute phase of bleeding. ADAMTS13 activity was measured at the same time, based on the FRETS-VWF73 assay [14] by SRL Co. (Tokyo, Japan). For the analysis, plasma was isolated in a platelet-free manner within 1 h after blood extraction and stored at −80°C until analysis. Quantification of large multimers of von Willebrand factor We first determined the concentrations of vWF in the plasma of the normal control subject and each patient by Western blot analysis under reduced conditions followed by densitometric analysis. The control plasma was prepared by extracting plasma from a single healthy person. A dispensed sample from the control plasma was consistently used for each time measurement throughout the entire study. The vWF antigen levels in the plasma of patient and control were determined by Western blot in a reduced condition. Then, plasma containing the same amount of vWF in each patient as that of the control was analysed in adjacent lanes by the vWF multimer analysis performed essentially as described [15] with polyclonal rabbit anti-human vWF antibody (DAKO, Glostrup, Denmark) as the primary antibody. The Western blots probing vWF were finally visualized by ImmunoStar® Zeta chemiluminescence (Wako, Osaka, Japan) and evaluated by Image Quant LAS 4000 mini (GE Healthcare) and ImageJ as described previously [13]. Bands 1–5 (starting from the lowest molecular weight band) were classified as low, 6–10 as medium and all those >10 as large multimers. The vWF large multimer proportion was calculated as the densitometric area of those bands. The vWF large multimer index was defined as the ratio of the large multimer proportion in total vWF derived from a patient to that from a normal control analysed in the adjacent lane [13]. Using this index, the quantity of vWF large multimers in a patient was expressed as a percentage of its normal control value obtained with simultaneous analysis of each time measurement (Fig. 1). Figure 1: View largeDownload slide Quantification method for calculation of the vWF large multimer index. vWF large multimer index in a patient is expressed as a percentage of the proportion of vWF large multimer in a normal control. LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 1: View largeDownload slide Quantification method for calculation of the vWF large multimer index. vWF large multimer index in a patient is expressed as a percentage of the proportion of vWF large multimer in a normal control. LVAD: left ventricular assist device; vWF: von Willebrand factor. Measurement of plasma von Willebrand factor antigen level and ristocetin cofactor activity vWF antigen (vWF: Ag) was measured by SRL Co. using the latex agglutination reaction with STA®-LIATEST VWF: Ag (Roche Diagnostics, Switzerland). vWF ristocetin cofactor activity (vWF: RCo) was also analysed by SRL Co. with BC von Willebrand Reagent (Siemens, Germany). Gastrointestinal bleeding events GIB was defined as any clinically suspected or documented suspicion of bleeding from the gastrointestinal tract as indicated with a de novo fall in haemoglobin level and the appearance of melena, haematochezia, haematemesis or guaiac positive stools as described previously [4, 16]. Statistical analyses All statistical analyses were performed using JMP® Pro version 13 (SAS Institute Inc., Cary, NC, USA). Continuous variables are expressed as mean ± standard deviation. Categorical variables are presented as frequency and percentages, and inter-group comparisons of categorical variables were analysed using Fisher’s exact test or χ2 test. If a value in a cell was <5, the Fisher’s exact test was applied. Intergroup comparison of continuous variables was performed using the Student’s t-test or the Welch’s t-test, if normally distributed; otherwise, Mann–Whitney U-test was used. Shapiro–Wilk test was applied to determine whether the variable follows a normal distribution. Kaplan–Meier methods were used to estimate GIB event-free survival. The median follow-up time was computed using the inverse Kaplan–Meier method. Receiver operating characteristic curves (ROC) were performed to identify the cut-off value of a vWF large multimer index. The difference was considered significant at a P-value <0.05. Ethics This study was performed in accordance with the Declaration of Helsinki. The study protocol was approved by the Institutional Ethics Committee in Tohoku University (Certification number; 2017-1-386). Written informed consent was obtained from each patient. RESULTS Patient characteristics Patient demographics and clinical information categorized by the 2 different pump types are summarized in Table 1. All patients were followed for a median of 591 days, and 28 patients were supported with LVADs throughout the follow-up periods. The remaining 13 patients were followed until the occurrence of the following events: heart transplantation in 5 patients on Days 591, 999, 1156, 1296 and 1550 after LVAD implantation; LVAD explantation in 1 patient due to severe device infection on Day 834; pump exchange in 4 patients on Days 384, 681, 778 and 434; and death in 3 patients on Days 248, 892 and 1361, respectively. No significant differences were observed in age, sex or aetiology between patients treated with axial (n = 33) and centrifugal (n = 8) types of LVADs (Table 1). Table 1: Characteristics of patients implanted with axial- and centrifugal-type LVADs LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; LVAD: left ventricular assist device; rpm: revolutions per minute; SD: standard deviation. Table 1: Characteristics of patients implanted with axial- and centrifugal-type LVADs LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    LVAD pump type  Axial (n = 33)  Centrifugal (n = 8)  P-value  Age (years), mean ± SD  44.5 ± 14.5  39.9 ± 9.75  0.29  Sex (male)  23  6  1.0  Aetiology      0.93   DCM  22  7     dHCM  4  0     ICM  3  1     Myocarditis  1  0     Congenital disease  2  0     Sarcoidosis  1  0    Device         HeartMate II  26       Jarvik 2000  7       Duraheart    6     EVAHeart    1     HVAD    1    State   Ongoing support  26  2     Death  2  1     Transplanted  2  3     Explanted  0  1     Pump exchanged  3  1    Complications  16  6     Thrombosis  5  2     Severe device infection  1  2     GIB  10  2    Follow-up periods after implantation (days)  478  1132    DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; LVAD: left ventricular assist device; rpm: revolutions per minute; SD: standard deviation. Gastrointestinal bleeding complications While thrombotic events of cerebral infarction, superior mesenteric artery thrombosis and pump thrombosis occurred in 7 patients and severe device infection occurred in 3 patients during the follow-up periods, GIB complications occurred in 12 (29%) patients at the median LVAD support time of 179 days (range 14–1106 days) (Tables 1 and 2). Ten GIB patients were supported with axial-type LVADs, and the other 2 with centrifugal-type LVADs. Mean haemoglobin levels in the acute bleeding phase of the 12 patients were 77.9 ± 17.1 g/l, suggesting that relatively large amounts of blood were lost on the bleeding events, whereas their PT-INR and APTT were within therapeutic ranges. The GIB-free survival curve indicated a 6-month event-free survival of 85.3%, 1-year survival of 77.0% and 2-year survival of 73.0% by the Kaplan–Meier method (Fig. 2). Four patients (Cases 1–4, Table 2) developed GIB in the early periods of 14–40 days after LVAD implantation. In 2 of these patients, bleeding lesions were detected by endoscopy. The bleeding sites contained oozing haemorrhages from ulcers and/or erosions, and angiodysplasia was not detected in either patient. On the other hand, GIB events occurred in the remaining 8 patients (Cases 5–12, Table 2) in the later periods of Day 142 and after. Upper and/or lower gastrointestinal examinations did not identify distinct bleeding lesions except for 2 patients. The origin of both bleeds in Case 5 at Day 142 and in Case 9 at Day 266 was confirmed to be angiodysplasia. Table 2: Clinical characteristics and course of 12 patients with GIB   Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected    Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected  a The data were examined in the acute phase of bleeding. b vWF multimer analysis and measurement of activity of ADAMTS13 were performed with patients in the stable state. c Case 3 was presented previously [17]. ADAMTS13: a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; Hb: haemoglobin; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; rpm: revolutions per minute; vWF: von Willebrand factor; vWF: RCo/vWF: Ag: ristocetin cofactor activity: von Willebrand factor antigen ratio. Table 2: Clinical characteristics and course of 12 patients with GIB   Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected    Sex  Age (years)  Aetiology  Device  Rotation speed (rpm)  Supported time until first bleeding (days)  Hb (g/l)a  vWF large multimer index (%)b  vWF: RCo/ vWF: Agb  ADAMTS13 activity (%)b  PT-INRa  Blood transfusion  Recurrence of GIB  Cause of GIB  Case 1  Male  63  ICM  HeartMate II  8400  14  68  33.4  0.72  117  2.05  Yes  No  Ulcer/erosion  Case 2  Male  42  Congenital  Jarvik 2000  10 000  17  69  32.2  0.60  120  1.79  Yes  Yes  Not detected  Case 3c  Male  61  DCM  Jarvik 2000  11 000  34  69  10.5  0.47  51  2.51  Yes  Yes  Ulcer/erosion9  Case 4  Male  49  DCM  HeartMate II  8200  40  108  34.6  0.58  136  2.35  No  No  Not detected  Case 5  Male  63  DCM  HeartMate II  8600  142  64  26.6  0.62  54  2.37  Yes  No  Angiodysplasia  Case 6  Male  49  DCM  HeartMate II  8600  155  64  7.1  0.18  122  2.24  Yes  No  Not detected  Case 7  Male  60  dHCM  HeartMate II  8800  202  72  11.5  0.58  135  2.00  No  Yes  Not detected  Case 8  Male  43  DCM  HVAD  2960  206  82  28.1  0.33  129  2.34  No  Yes  Not detected  Case 9  Male  62  DCM  Jarvik 2000  9000  266  71  23.7  0.53  119  1.56  No  Yes  Angiodysplasia  Case 10  Female  23  dHCM  HeartMate II  8200  574  113  23.6  0.23  154  2.48  No  No  Not detected  Case 11  Male  33  DCM  Duraheart  1600  836  64  39.4  0.10  140  2.70  Yes  No  Not detected  Case 12  Male  40  DCM  HeartMate II  8400  1106  91  29.5  0.11  170  1.98  No  Yes  Not detected  a The data were examined in the acute phase of bleeding. b vWF multimer analysis and measurement of activity of ADAMTS13 were performed with patients in the stable state. c Case 3 was presented previously [17]. ADAMTS13: a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; Hb: haemoglobin; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; rpm: revolutions per minute; vWF: von Willebrand factor; vWF: RCo/vWF: Ag: ristocetin cofactor activity: von Willebrand factor antigen ratio. Figure 2: View largeDownload slide Actuarial freedom from first gastrointestinal bleeding after continuous-flow type left ventricular assist device implantation. GIB: gastrointestinal bleeding. Figure 2: View largeDownload slide Actuarial freedom from first gastrointestinal bleeding after continuous-flow type left ventricular assist device implantation. GIB: gastrointestinal bleeding. The characteristics of the LVAD patients with or without GIB are summarized in Table 3. No significant differences were observed in gender, age, aetiology of heart disease, pump types or haematological data between the 2 groups using the Fisher’s exact test. All the other laboratory findings were not significantly different, except for creatinine (serum creatinine levels: GIB groups 1.29 ± 0.68 vs Non-GIB groups 0.74 ± 0.31 mg/dl, P = 0.017) by the Student’s t-test or the Welch’s t-test, when appropriate. Table 3: Characteristics of LVAD Patients with or without GIB   Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48    Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48  APTT: activated partial thromboplastin time; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; SD: standard deviation. Table 3: Characteristics of LVAD Patients with or without GIB   Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48    Non-GIB (n = 29)  GIB (n = 12)  P-value  Age (years), mean ± SD  41.3 ± 13.4  49.0 ± 13.2  0.10  Sex (male)  18  11  0.07  Aetiology      0.90  DCM  21  8     dHCM  2  2     ICM  3  1     Myocarditis  1  0     Sarcoidosis  1  0     Congenital disease  1  1    Pump type (axial)  23  10  1.0  Device         HeartMate II  19  7     Jarvik 2000  4  3     Duraheart  5  1     EVAHeart  1  0     HVAD  0  1    Blood type (Type O)  11  4  1.0  Haematological data    (in stable phase)  (in bleeding phase)     PT-INR, mean ± SD  2.14 ± 0.63  2.20 ± 0.33  0.77   APTT (s), mean ± SD  44.5 ± 7.0  44.8 ± 5.6  0.87   Platelet count (×103/μl), mean ± SD  224 ± 91  203 ± 83  0.48  APTT: activated partial thromboplastin time; DCM: dilated cardiomyopathy; dHCM: dilated phase of hypertrophic cardiomyopathy; GIB: gastrointestinal bleeding; ICM: ischaemic cardiomyopathy; PT-INR: prothrombin time–international normalized ratio; SD: standard deviation. Evaluation of vWF large multimers by the vWF large multimer index All LVAD patients exhibited severe loss of vWF large multimers [18], with a mean index of 33.8 ± 16.8%. With regard to pump types, vWF large multimer indices in patients with axial-type LVADs were significantly lower compared to those with centrifugal-type LVADs (31.4 ± 16.7% vs 43.8 ± 14.2%, P = 0.030) by the Student’s t-test (Fig. 3). Those in patients with GIB were significantly lower compared to those without GIB (25.0 ± 10.3% vs 37.5 ± 17.8%, P = 0.008) by the Welch’s t-test (Fig. 4). Notably, all the patients with GIB exhibited indices below 40%, and no patients with an index above 40% developed GIB. Furthermore, 12 (41%) of the 29 patients with indices below 40% experienced GIB during the median follow-up period of 591 days. ROC identified the cut-off value for vWF large multimer on GIB as 39.4% (sensitivity 100%, specificity 41.4%) (see Supplementary Material, Fig. S1). On the other hand, the other parameters including vWF: Ag (103.6 ± 50.4% vs 116.4 ± 57.7%, P = 0.74), vWF: RCo (49.8 ± 39.0% vs 61.3 ± 44.2%, P = 0.78), vWF: RCo/vWF: Ag (0.52 ± 0.28 vs 0.42 ± 0.22, P = 0.85) and ADAMTS13 activities (120.6 ± 35.3% vs 125.5 ± 36.6%, P = 0.69) were not statistically different (GIB group vs Non-GIB group, respectively) by the Student’s t-test or the Welch’s t-test, when appropriate. Figure 3: View largeDownload slide Comparison of vWF large multimer indices in patients with axial-type LVADs (n = 33) and centrifugal-type LVADs (n = 8). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. Cases who developed gastrointestinal bleeding are circled in red. GIB: gastrointestinal bleeding; LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 3: View largeDownload slide Comparison of vWF large multimer indices in patients with axial-type LVADs (n = 33) and centrifugal-type LVADs (n = 8). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. Cases who developed gastrointestinal bleeding are circled in red. GIB: gastrointestinal bleeding; LVAD: left ventricular assist device; vWF: von Willebrand factor. Figure 4: View largeDownload slide Comparison of vWF large multimer indices in left ventricular assist device patients with (n = 12) and without GIB (n = 29). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. GIB: gastrointestinal bleeding; vWF: von Willebrand factor. Figure 4: View largeDownload slide Comparison of vWF large multimer indices in left ventricular assist device patients with (n = 12) and without GIB (n = 29). The black dots indicate the measured values from each patient, and the upper and lower error bars indicate the standard deviation, and the middle bar is the mean value. GIB: gastrointestinal bleeding; vWF: von Willebrand factor. DISCUSSION In this study, we have quantitatively evaluated the loss of vWF large multimers that could indicate the haematological severity of aVWS in 41 patients after an LVAD implantation. We have demonstrated that GIB occurred in 12 (29%) patients who exhibited a more severe loss of vWF large multimers than those without GIB. Further, cut-off value for vWF large multimer index on GIB was 39.4% by ROC analysis. Thus, aVWS in a severe form associated with a certain threshold of multimer index may dictate the risk of LVAD-associated GIB during the course of follow-up. In some previous studies, vWF large multimers were evaluated by vWF large multimer ratios [13, 19, 20]. In a study by Tamura et al. [13] analysing 31 severe aortic stenosis patients, vWF large multimer ratios were decreased in a pressure gradient-dependent manner. Their study demonstrated that the ratios were distributed predominantly between 10% and 35%, although these ratios overlapped extensively with those of normal controls analysed simultaneously [13]. In another study, vWF large multimer ratios have also been shown to decrease depending on the severity of aortic stenosis where by ratios were distributed between 4% and 14% [19]. Furthermore, Meyer et al. [20] demonstrated that in 102 patients treated with LVAD, pump speed was weakly, but significantly, correlated inversely with vWF large multimer ratios distributed between 20% and 60%. It has been reported that vWF multimer profiles do not correlate with the incidence of bleeding complications [20]. Thus, although these ratios may indicate some association with clinical parameters, the absolute values of the vWF large multimer ratios vary considerably between studies. Therefore, it would be difficult to use the ratios for comparison across studies. ROC vWF large multimer analysis is categorized as a manoeuvre of Western blotting [15], the results obtained would appear to be influenced by the vWF antigen amount to be analysed and the exposure time in the final visualization procedure. Variability of vWF large multimer ratios may be attributable to those factors associated with laboratory work. To avoid such variability, we have recently proposed a novel value, namely the vWF large multimer index [13], since the index is a relative value using vWF large multimers of a normal control as a reference. We further confirmed using our quantitative method that the axial pump group exhibited more severe loss of vWF large multimers than the centrifugal pump group as shown previously [20, 21]. It is notable that 39.4% could be a predictive value for subsequent GIB in patients treated with continuous-flow-type LVADs, irrespective of either axial- or centrifugal-type pump involvement. This index may also be useful in avoiding GIB complications by titrating the intensity of antithrombotic therapy, since vWF is a platelet-binding factor and plays a critical role in haemostasis. It may be worthwhile to consider modifying an the antithrombotic regimen for patients developing severe aVWS, who exhibit a vWF large multimer index lower than 40%. However, since the number of patients in this study was small, a modification of the anticoagulation and antiplatelet therapy protocol in patients with LVAD-associated aVWS warrants further larger scale clinical investigation to establish conclusive evidence. The causes of GIB are clinically considered to be multifactorial. Having severe aVWS as a substrate, development of gastrointestinal ulcer at an early stage and angiodysplasia at a later stage may evoke GIB as a superimposed factor in patients treated with LVAD. In our present study, the majority of GIB patients did not have a clear cause of GIB, and only 4 patients were clearly found to have bleeding lesions such as ulcer, erosion and angiodysplasia. Among those 4 patients the mean value of the vWF multimer index was 25.2%, whereas the patients without clear causes of GIB exhibited a mean index of 28.8%. There were no significant differences (P = 0.67) in those values. The correlation between the vWF large multimer index and development of gastrointestinal organic lesion was not prove, so further study will be required to clarify this issue. Limitations The limitations to this study include the fact that it was a retrospective single-centre study with a small number of patients. Further study is required for unequivocal establishment of the vWF large multimer index as a predictive measure for GIB complications in patients implanted with LVADs. CONCLUSIONS The vWF large multimer index, which is a novel quantitative methodology employed to evaluate the haematological severity of aVWS, was applied in all 41 patients after LVAD implantation. Using this index, patients with GIB exhibited more severe aVWS compared to those without GIB haematologically. GIB patients exhibited severe aVWS with large multimer indices of approximately ≤40%. The vWF large multimer index may predict a risk of GIB after an LVAD implantation. SUPPLEMENTARY MATERIAL Supplementary material is available at EJCTS online. FUNDING This work was supported by JSPS KAKENHI [15K10204 to S.K., M.A. and Y.S.] and partly supported by a Health and Labor Sciences Research Grant for Research on rare and intractable diseases from the Ministry of Health, Labor and Welfare, Japan, to H.H. and Y.S., and grants from the SENSHIN Medical Research Foundation and the Suzuken Memorial Foundation to H.H. Conflict of interest: none declared. REFERENCES 1 Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W et al.   Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med  2001; 345: 1435– 43. Google Scholar CrossRef Search ADS PubMed  2 Hata H, Fujita T, Shimahara Y, Sato S, Yanase M, Seguchi O et al.   Early and mid-term outcomes of left ventricular assist device implantation and future prospects. Gen Thorac Cardiovasc Surg  2015; 63: 557– 64. Google Scholar CrossRef Search ADS PubMed  3 Suarez J, Patel CB, Felker GM, Becker R, Hernandez AF, Rogers JG. Mechanisms of bleeding and approach to patients with axial-flow left ventricular assist devices. Circ Heart Fail  2011; 4: 779– 84. Google Scholar CrossRef Search ADS PubMed  4 Draper KV, Huang RJ, Gerson LB. GI bleeding in patients with continuous-flow left ventricular assist devices: a systematic review and meta-analysis. Gastrointest Endosc  2014; 80: 435– 46.e1. Google Scholar CrossRef Search ADS PubMed  5 Demirozu ZT, Radovancevic R, Hochman LF, Gregoric ID, Letsou GV, Kar B et al.   Arteriovenous malformation and gastrointestinal bleeding in patients with the HeartMate II left ventricular assist device. J Heart Lung Transplant  2011; 30: 849– 53. Google Scholar CrossRef Search ADS PubMed  6 Uriel N, Pak SW, Jorde UP, Jude B, Susen S, Vincentelli A et al.   Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation. J Am Coll Cardiol  2010; 56: 1207– 13. Google Scholar CrossRef Search ADS PubMed  7 Geisen U, Heilmann C, Beyersdorf F, Benk C, Berchtold-Herz M, Schlensak C et al.   Non-surgical bleeding in patients with ventricular assist devices could be explained by acquired von Willebrand disease. Eur J Cardiothorac Surg  2008; 33: 679– 84. Google Scholar CrossRef Search ADS PubMed  8 Shida Y, Nishio K, Sugimoto M, Mizuno T, Hamada M, Kato S et al.   Functional imaging of shear-dependent activity of ADAMTS13 in regulating mural thrombus growth under whole blood flow conditions. Blood  2008; 111: 1295– 8. Google Scholar CrossRef Search ADS PubMed  9 Matsumoto M, Kawaguchi S, Ishizashi H, Yagi H, Iida J, Sakaki T et al.   Platelets treated with ticlopidine are less reactive to unusually large von Willebrand factor multimers than are those treated with aspirin under high shear stress. Pathophysiol Haemost Thromb  2005; 34: 35– 40. Google Scholar CrossRef Search ADS PubMed  10 Malehsa D, Meyer AL, Bara C, Struber M. Acquired von Willebrand syndrome after exchange of the HeartMate XVE to the HeartMate II ventricular assist device. Eur J Cardiothorac Surg  2009; 35: 1091– 3. Google Scholar CrossRef Search ADS PubMed  11 Nascimbene A, Neelamegham S, Frazier OH, Moake JL, Dong JF. Acquired von Willebrand syndrome associated with left ventricular assist device. Blood  2016; 127: 3133– 41. Google Scholar CrossRef Search ADS PubMed  12 Baghai M, Heilmann C, Beyersdorf F, Nakamura L, Geisen U, Olschewski M et al.   Platelet dysfunction and acquired von Willebrand syndrome in patients with left ventricular assist devices. Eur J Cardiothorac Surg  2015; 48: 421– 7. Google Scholar CrossRef Search ADS PubMed  13 Tamura T, Horiuchi H, Imai M, Tada T, Shiomi H, Kuroda M et al.   Unexpectedly high prevalence of acquired von Willebrand syndrome in patients with severe aortic stenosis as evaluated with a novel large multimer index. J Atheroscler Thromb  2015; 22: 1115– 23. Google Scholar CrossRef Search ADS PubMed  14 Kokame K, Nobe Y, Kokubo Y, Okayama A, Miyata T. FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol  2005; 129: 93– 100. Google Scholar CrossRef Search ADS PubMed  15 Ruggeri ZM, Zimmerman TS. The complex multimeric composition of factor VIII/von Willebrand factor. Blood  1981; 57: 1140– 3. Google Scholar PubMed  16 Goldstein DJ, Aaronson KD, Tatooles AJ, Silvestry SC, Jeevanandam V, Gordon R et al.   Gastrointestinal bleeding in recipients of the HeartWare Ventricular Assist System. JACC Heart Fail  2015; 3: 303– 13. Google Scholar CrossRef Search ADS PubMed  17 Sakatsume K, Akiyama M, Saito K, Kawamoto S, Horiuchi H, Saiki Y. Intractable bleeding tendency due to acquired von Willebrand syndrome after Jarvik 2000 implant. J Artif Organs  2016; 19: 289– 92. Google Scholar CrossRef Search ADS PubMed  18 Heilmann C, Geisen U, Beyersdorf F, Nakamura L, Trummer G, Berchtold-Herz M et al.   Acquired Von Willebrand syndrome is an early-onset problem in ventricular assist device patients. Eur J Cardiothorac Surg  2011; 40: 1328– 33; discussion 233. Google Scholar PubMed  19 Vincentelli A, Susen S, Le Tourneau T, Six I, Fabre O, Juthier F et al.   Acquired von Willebrand syndrome in aortic stenosis. N Engl J Med  2003; 349: 343– 9. Google Scholar CrossRef Search ADS PubMed  20 Meyer AL, Malehsa D, Budde U, Bara C, Haverich A, Strueber M. Acquired von Willebrand syndrome in patients with a centrifugal or axial continuous flow left ventricular assist device. JACC Heart Fail  2014; 2: 141– 5. Google Scholar CrossRef Search ADS PubMed  21 Netuka I, Kvasnicka T, Kvasnicka J, Hrachovinova I, Ivak P, Marecek F et al.   Evaluation of von Willebrand factor with a fully magnetically levitated centrifugal continuous-flow left ventricular assist device in advanced heart failure. J Heart Lung Transplant  2016; 35: 860– 7. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: May 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off