Ann Thorac Surg 2002;73:546-548
© 2002 The Society of Thoracic Surgeons
Original article: cardiovascular
And hemolysis goes on: ventricular assist device in combination with veno-venous hemofiltration
Heyman Luckraz, FRCS*a,
Michael Woods, RGNa,
Stephen R. Large, FRCSa The Papworth VAD Group,a
a The Transplant Unit, Papworth Hospital, Papworth Everard, Cambridgeshire, United Kingdom
Accepted for publication September 13, 2001.
* Address reprint requests to Mr Luckraz, The Transplant Unit, Papworth Hospital, Papworth Everard, Cambridgeshire CB3 8RE, UK
e-mail: heyman.luckraz{at}papworth-tr.anglox.nhs.uk
 |
Abstract
|
|---|
Background. Ventricular Assist Device (VAD) is an accepted treatment as a bridge to cardiac transplantation, and may be of help in patients as destination therapy for end-stage cardiac failure. The low output state associated with end-stage cardiac failure predisposes patients to renal dysfunction and the need for short-term renal support. The use of cardiopulmonary bypass for VAD insertion, VAD, and hemofiltration expose the blood to mechanical trauma and activated inflammatory cascades that can result in hemolysis. This produces free hemoglobin, a known nephrotoxin; this is a further renal insult. This study assesses the effect of VAD alone and in combination with continuous veno-venous hemofiltration (CVVHF) on hemolysis.
Methods and Results. From July 1999 to December 2000, Thoratec VAD was used in 11 patients. Nine (all males) were included in this study as all had laboratory profiles. Hemolysis was quantified by plasma free hemoglobin (PFHb) and hydroxybuterate dehydrogenase (HBD) levels measured daily, defined as PFHb level greater than 40 mg/L and HBD greater than 250 IU/L. Data relate to the following time intervals while the VAD was still in situ: T1 = 24 hours post-VAD insertion, T2 = 24 hours post-CVVHF start, T3 = 48 to 72 hours with the same CVVHF circuit, T4 = 24 hours post-stopping of CVVHF, and T5 = CVVHF off for over 48 hours. The mean (SD) PFHb levels were 19.6 (10.9) at T1, 31.7 (0.6) at T2, 93.7 (16.4) at T3 (p < 0.05), 32.5 (20.9) at T4, and 14.2 (3.8) at T5 (p < 0.05). These changes were paralleled by the mean (SD) HBD levels: T1 = 1,337 (616), T2 = 2,025 (509), T3 = 2,676 (1,170) (p < 0.05), T4 1,780 (618), and T5 = 1,310 (436).
Conclusions. Thoratec VAD was associated with a mild degree of hemolysis. This was worsened by concomitant use of CVVHF. The effect was accentuated if the same CVVHF circuit was used for over 48 hours but was reversible within 24 hours of stopping the hemofilter.
|
Introduction
|
|---|
Ventricular Assist Devices (VAD) are indicated in cardiac failure as a bridge to either heart transplantation or rarely, myocardial recovery. The low output state, which precedes VAD insertion, coupled with cardiopulmonary bypass expose patients to acute renal insult. It is recognized that 54% of patients undergoing a Thoratec VAD (Thoratec Laboratories Corp, Pleasanton, CA) experience renal dysfunction as defined by a rise of 1.5 times the upper limit of normal of creatinine [1]. This mechanical device is also responsible for hemolysis and the increased production of plasma free hemoglobin (PFHb) three times the normal limits in up to 51% of patients [1]. PFHb is nephrotoxic and worsens the already compromised renal function so that short-term renal support is mandatory. The latter may be provided by using continuous veno-venous hemofiltration (CVVHF). However, the extracorporeal circuit for CVVHF adds further mechanical trauma to the blood cells and also activates inflammatory cascades with an end result of increased hemolysis.
We describe the magnitude of hemolysis when hemofiltration is added to mechanical heart support. It is our belief that it is significant when CVVHF is combined with VAD. This increases the risk of renal failure.
|
Material and methods
|
|---|
The Thoratec VAD was used in all 9 patients. Vascular access for CVVHF was by a dual lumen dialysis catheter, 12 Fr (Vygon, GmbH & Co, Aachen, Germany), which was inserted either in the subclavian or internal jugular vein using the Seldinger technique. The hemofiltration roller pump was the Hospal BSM 22, with the infusion lines ("arterial" and "venous") 8 and 3.9 mm diameter, respectively (Hospal DASCO, Medolla, Italy). A membrane hollow fiber filter with a wet surface area of 1.3 m2 with a pump flow 150 to 200 mL/h was used. PFHb levels were measured by spectrophotometry.
|
Results
|
|---|
From July 1999 to December 2000, 11 patients had a Thoratec VAD device inserted at our institution. Two patients (both females) were excluded due to incomplete data. The remaining 9 patients (all males) were followed through their postoperative phase with a daily hemolytic screen until VAD explantation. Data are expressed either as percentages (%) or mean (SD). Three patients (33%) did not require hemofiltration following VAD insertion while the remaining 6 patients were hemofiltered for 12 days. The activated clotting time during CVVHF was maintained between 180 and 200 seconds with a continuous heparin infusion. The mean age of the whole group was 37.8 (15.6) years. There were two deaths (one due to sepsis and the other due to ischemic bowel at 8 and 28 days, respectively). One of the remaining 7 patients had the device explanted following complete myocardial recovery and the remainder (6 patients) have had successful heart transplantation.
The levels of hydroxybuterate dehydrogenase (HBD) and PFHb were measured daily following VAD insertion. The blood samples were collected by gentle aspiration from the arterial cannulation line after discarding of the initial 10 mL. The levels of PFHb and HBD along with their corresponding venous line pressure of the CVVHF are shown in Table 1.
There was significant increase in the PFHb and HBD levels when the same filter circuit was used for over 48 hours. This was most likely due to micro-clot formation in the circuit as evidenced by the gradual increase in venous pressure line, which became significant when the circuit life was over 48 hours. Overall, six BiVADs and three LVADs (66% of whom required CVVHF) were implanted. The PFHb level for the BiVAD group on CVVHF increased from 25 (17) mg/L (prehemofilter) to 103 (4) mg/L (48 hours post-start of CVVHF) (p < 0.05). The respective values for the LVAD/CVVHF group were 21 and 75 mg/L. The creatinine levels on admission for those patients requiring hemofiltration were 212 (92) µmol/L, rising to 240 (116) µmol/L 24 hours after VAD insertion, falling to 168 (40) µmol/L while on the hemofilter and stabilizing at 104 (20) µmol/L after VAD removal (Fig 1).
View larger version (23K):
[in this window]
[in a new window]
|
Fig 1. Box and whisker plot illustrating the creatinine levels in µmol/L (y-axis) at different time scales: admission to hospital, pre-VAD (A); 24 hour post-VAD insertion (B); VAD in situ, while on hemofilter (C); and VAD explanted, prior to hospital discharge (D). (VAD = ventricular assist device.)
|
|
|
Comment
|
|---|
Rise in PFHb levels during Thoratec VAD usage has previously been described by Farrar and Hill [2]. These changes were reported as transient and usually settling within 2 weeks of VAD insertion. Latest data from Thoratec [1] suggest that hemolysis is experienced in up to 51% of VAD patients (without concomitant use of CVVHF) as judged by a threefold increase of the PFHb levels. In extracorporeal membrane oxygenator circuits, PFHb levels of up to 40 mg/L reflect nonsignificant hemolysis [3].
Similar hemolytic effects of CVVHF have been documented by Bierer and colleagues [4]. They confirmed that the hemolytic screen post-CVVHF showed significant (p < 0.01) increase in PFHb levels but only when the same CVVHF circuit is used for over 50 hours. This was associated with an increase in the venous line pressure, implying clot formation within the circuit. As the CVVHF is pump-driven, it takes several days for the clots to impair the filtration process. In order to preempt this problem, Macias and associates suggested changing the hemofilter circuit every 48 hours [5]. However, this was thought not to be cost-effective and, hence, circuits continue to be used for over 48 hours [4]. It has also been suggested that hemolysis is related to the size of the vascular access cannula [6]. This was documented by measuring the modified index of hemolysis of various cannula diameters: the smaller the cannula size, the greater the modified index of hemolysis. This is not too surprising as the blood is exposed to the mechanical trauma twice through the cannula, once at outflow to the filter circuit and again on "arterial" return to the patient.
In our series, the usage of CVVHF concurrently with the Thoratec VAD was associated with a significant increase in hemolysis as measured by PFHb and HBD. As PFHb is nephrotoxic, renal impairment is worsened thus prolonging the need for renal support. This not only increases the duration of hospital stay, but persistent renal failure significantly affects outcome following cardiac transplantation. There is lengthened anxiety for the patient and increased hospital costs.
Although this report included only 9 patients, there is convincing evidence of significant hemolysis occurring when CVVHF is used in combination with VADs, especially if the CVVHF circuit is over 48 hours. Hence, there may be some benefit in limiting the circuit life to 48 hours in the group of patients who already have a VAD in situ. Perhaps it would be appropriate to use peritoneal dialysis to get around that problem. Furthermore, consideration should be given to using vascular cannulae, which are less likely to promote further red cell destruction. This can be achieved by using a larger bore and shorter length cannula. Hemolysis will thus be reduced as will renal insult due to a high level of PFHb.
In conclusion, Thoratec VAD was associated with a mild degree of hemolysis worsened by concomitant use of CVVHF. This may be improved by limiting the lifespan of the CVVHF circuit, altering the circuit design, or using alternatives to CVVHF such as peritoneal dialysis.
|
References
|
|---|
-
Thoratec Ventricular Assist Device System. Directions for use. Thoratec Laborotories Corporation, Pleasanton, California. September 1999:5.
-
Farrar D.J., Hill J.D. Univentricular and biventricular Thoratec VAD support as a bridge to transplantation. Ann Thorac Surg 1993;55:276-282.[Abstract]
-
Funakubo A., Sakuma I., Fului Y., Kawamura T. Development of a compact extracorporeal membrane oxygenation system. Artif Organs 1991;15:56-59.[Medline]
-
Bierer P., Holt A.W., Bersten D., Plummer J.L., Chalmers A.H. Haemolysis associated with continuous venovenous renal replacement circuits. Anaesth Intensive Care 1998;26:272-275.[Medline]
-
Macias W.L., Mueller B.A., Scarim S.K., Robinson M., Rudy D.W. Continuous endogenous haemofiltration: an alternative to continuous arteriovenous haemofiltration and haemodiafiltration in acute renal failure. Am J Kidney Dis 1991;18:451-458.[Medline]
-
De Wachter D.S., Verdonck P.R., De Vos V.Y., Hombrouckx R.O. Blood trauma in plastic haemodialysis cannulae. Int J Artif Organs 1997;20:366-370.[Medline]
This article has been cited by other articles:
|
|
|
|
 
L. E. Samuels, E. C. Holmes, P. Garwood, and F. Ferdinand
Initial Experience With the Abiomed AB5000 Ventricular Assist Device System
Ann. Thorac. Surg.,
July 1, 2005;
80(1):
309 - 312.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
