Abstract 1122


Continuous flow left ventricular assist device (CF-LVAD) recipients have increased tendency of post-surgical bleeding. We have previously described a universal loss of high molecular weight VWF multimers (HMWM), likely due to enhanced cleavage of the HMWM in CF-LVAD recipients. Among various VWF laboratory tests, VWF multimer analysis remains the reference method for detecting such an acquired VWF abnormality (AVWA). We hypothesize that the severity of AVWA is a predictor for bleeding tendency. In this study, using our in-gel VWF multimer electrophoresis method, we measured the severity of HMWM loss in CF-LVAD recipients by infra-red densitometry analysis and examined its association with other VWF activity test results and patients' bleeding tendency.


As part of an on-going prospective multicenter study, pre- and/or post-CF-LVAD implantation (7, 30 days and 5–7 months) blood samples were collected from 20 LVAD recipients during 2008 and 2009. Their clinically significant bleeding episodes were recorded. Plasma VWF antigen (VWF:Ag), VWF:Lx activity [Chen D, et al. J Thromb Haemost. 2011;9(10):1993–2002] and plasma VWF multimer analyses [Pruthi RK, et al. Thromb Res. 2010;126(6):543–9] were performed on all available plasma samples. VWF multimers were quantified by infra-red fluorescent densitometry analysis using Odyssey Infrared Imager version 2.0. VWF:multi ratio defined by density ratio between the high molecular weight multimer (HMWM, bands 11 and above) and low and intermediate molecular weight multimer (band 2 to 9) was calculated. Normal range of VWF:multi ratio was determined from 16 normal donor plasma and 4 pooled normal plasma. Statistical analyses including t-test and kappa test were employed.


The normal range of VWF:multi ratio is 0.4–0.76. The CV of VWF:multi ratio of a pooled normal plasma tested in 21 gel runs is 12%. Loss of VWF HMWM is present in three patients before CF-LVAD implantation and all patients after implantation. All samples with loss of HMWM show decreased VWF:multi ratio (<0.4). At a cut-off value of 0.4, the sensitivity and specificity of VWF:multi ratio for detecting HMWM loss are 100% and 95.5% respectively (n=67). Decrease of VWF:multi ratio is observed after implantation (0.41±0.07 pre- vs. 0.29±0.07 post-implantation, P<0.00001). In four patients who received heart transplantation, normalization of VWF multimer pattern and VWF:multi ratio (0.54±0.06) is observed. Ten patients had significant post-implantation bleeding episodes that are associated with significantly decreased VWF:multi ratio (0.26±0.04 bleeding vs. 0.32±0.09 non-bleeding, P=0.002) and VWF Lx/Ag ratio (0.72±0.10 bleeding vs 0.80±0.09 non-bleeding, P=0.001). At a cut-off of 0.26, VWF:multi ratio demonstrates a sensitivity of 64% and specificity of 85.7% for predicting clinically significant bleeding. The VWF:multi ratio shows moderate agreement with VWF:Lx/Ag ratio (Kappa value 0.42).


Severity of VWF HMWM losses can be quantitatively measured by the new in-gel infra-red densitometry method (VWF:multi ratio). Severity of VWF HMWM losses defined by the VWF:multi ratios, along with VWF:Lx/Ag ratios are significantly associated with bleeding tendency in LVAD recipients.


Milano:Thoratec: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.