INTRODUCTION

There is increasing concern about the negative side effects of allogeneic blood transfusion on post-operative outcome [1-2-3-4]. This has promoted the utilization of new strategies to decrease the use of banked blood products in pediatric cardiac surgery. In this regard, reduction of cardiopulmonary bypass prime volume has proved to be effective [5-6-7-8-9]. Downsizing of the by pass circuit was possible because the manufacturers currently provide components with low prime volume: 43 ml for the Baby-FX 05 with an integrated arterial filter or 45 ml for the Kids D100 with an arterial filter and 31 ml without an arterial filter. Thus, the perfusionist is now responsible for the success of this operation.

Three main ways are described to optimize the prime volume decrease:
1 - A circuit without an arterial filter and without a hemofilter. These two components currently used in pediatric centers, are not essential and have already been withdrawn in several centers without obvious drawbacks [7-10-11-12]
2 - A two-level heart-lung machine with arterial and cardioplegia roller pumps at the lower level and suction roller pumps at the upper level. The upper position of the suction roller pumps allows reduction in the length of the left ventricular vent and of the cardiotomy suction lines [Fig. 1].
3 - The use of vacuum-assisted venous return with limited gravity siphon-dependant drainage, with the top level of the cardiotomy reservoir at about the level of the patient right atrium. The upper position of the membrane oxygenator allows reduction in the length of the arterial and venous lines, as well as a further reduction in the length of the suction lines (with the venous line, the left vent and the suction lines being in an horizontal position) [Fig. 2].

Fig. 1. The two-level heart-lung machine.


Fig. 2. The venous line. The top of the cardiotomy
reservoir is at about the level of the right atrium.

VACUUM-ASSISTED VENOUS DRAINAGE

Vacuum-assisted venous drainage (VAVD) was used in the early years of cardiac surgery [13] and then abandoned until a renewal of interest, for non-routine indications in the 90s. At that time, the main indications were minimally invasive surgery or redo-operations with significant sternal adhesions, which were performed using extra thoracic cardiopulmonary bypass. Femoral venous siphon-gravity drainage was often inadequate in providing full-flow support and augmented venous return techniques were necessary to optimize venous return. VAVD proved to be efficient in performing full-flow bypass with small tubing and cannulas, but specific drawbacks were also described.

Air embolism

During VAVD there are two opposite factors that influence the value of the pressure in the cardiotomy reservoir. The vacuum regulator is set to adjust a negative pressure in the hardshell reservoir, while suction rotor pumps, introducing air and blood in the cardiotomy reservoir, tend to counteract this negative pressure.

An inadvertent positive pressurization of the reservoir may stop venous drainage and force air in the venous line, resulting in right atrial air embolus. Therefore, if a patent atrial defect is present, a paradoxic air embolus may occur [14]. The majority of recent oxygenators have built-in security pressure valves which open in the case of positive pressure. When using a membrane oxygenator without built-in valve, a separate disposable valve must be connected to one port of the rigid reservoir.

When a non-occlusive pump, like a centrifugal pump, is used as an arterial pump, the negative pressure is transmitted from the cardiotomy reservoir to the membrane oxygenator. In vitro tests were performed by Jegger et al. on six different oxygenators to evaluate the risk of VAVD associated with a non-occlusive pump [15]. With a centrifugal pump placed between the cardiotomy reservoir and the membrane oxygenator, a mean negative pressure of - 67 +/- 7 mm Hg was necessary to generate bubbles in the blood compartment of the oxygenator. Furthermore, when the negative pressure values exceed 60 mmHg, the transmission of the negative pressure to the centrifugal pump is likely to induce a decrease in real blood flow. Fiorucci et al. described an overestimation of blood flow as great as 54 % of the measured flow when a 100 mm Hg negative pressure is applied upstream of the centrifugal pump [16].

There has also been concern about the risk of increasing gaseous microemboli delivered to the patient when using VAVD. An experimental study performed by Jones et al. concluded that VAVD, using a negative pressure of up to 40 mmHg, did not significantly increase gaseous microemboli, when compared to gravity-siphon venous drainage [17]. Most of the studies describing an exacerbation of arterial line microemboli during VAVD were performed with higher negative pressure (60 or 80 mm Hg).

Blood trauma

Another concern during VAVD was the risk of increased damage to blood cells. This drawback was ruled out in an experimental animal study from Mueller et al. [18] and in an in vitro study by Mathews et al. [19]. The absence of any significant impact on hemolysis was confirmed later in a human randomized prospective study [20].

Wall vacuum failure

Whenever the full-flow of the bypass depends on VAVD, security is needed in case of wall vacuum failure. In case of an unexpected wall vacuum failure, Hill has described the efficiency of an occlusive roller pump (with tubing connected to a sucker port of the reservoir) in generating a negative pressure in the hard-shell reservoir [21].

Most of the complications related to VAVD arose during its "learning curve". However, the advantages of VAVD also became evident; therefore, its application was broadened to include all open-heart procedures.

Before deciding to move to routine use of VAVD, and also to gain familiarity with the technique, we have performed "in vitro" tests in the laboratory. We have confirmed the two major advantages of VAVD during this experimental phase:

- The downsizing of bypass circuits with smaller prime volume and smaller volume of blood diverted in the suction lines, when compared to gravity-siphon venous drainage.
- The efficiency of small negative pressure variation in inducing large increases in blood drainage (Fig. 3).


Fig. 3. Variations of blood drainage (hematocrit = 24%) in 3/16 inch silicone tubing. The maximal siphon-gravity venous drainage was 160 ml, with a declivity of 6 cm. The maximal venous drainage was 1340 ml when adding a negative pressure of -45 mmHg.

Methods and Results

Since February 2007, the VAVD technique has been used routinely in our unit. Every effort was made to decrease prime volume, namely:

- the use of a dedicated pediatric heart lung machine with two levels, the simplest bypass circuit without an arterial filter or a hemofilter
- a high position of the membrane oxygenator (Fig. 1).

We used a vacuum regulator (Precision Medical, Northampton, PA) connected to wall vacuum. The pressure in the cardiotomy reservoir was continuously monitored with a pressure transducer and displayed on the screen of the heart-lung machine (Stoëckert S5, Munchen, Germany).

More than 800 procedures were performed with this technique without obvious drawback. The major advantages of VAVD were expected for the neonates, infants and small children. Two different bypass circuits with two different prime volumes, including the prime volume of the cardioplegia circuit, were used:

- For neonates and infants with bypass flow up to 1000 ml, arterial and venous lines were composed of 3/16 inch tubing, and the prime volume was 120 ml. The level in the membrane oxygenator was about 40 ml prior to initiation of bypass (i.e., after elimination of the shunt between the arterial and the venous line, connection to the arterial and venous cannulas and salvage of the blood content of the eliminated tubing).
- For infants and young children with bypass flow between 1000 ml and 1500 ml, the venous line was 1/4 inch tubing and the arterial line was 3/16 inch tubing. The prime volume was170 ml and the level in the membrane oxygenator was about 50 ml prior to initiation of bypass.

VAVD was always needed to obtain full-flow support. The mean negative pressure in the cardiotomy reservoir was between - 20 and - 30 mm Hg in a vast majority of cases. In rare instances, a higher negative pressure is necessary during short periods, but we never exceed - 40 mm Hg. We try to maintain the negative pressure as low as possible; therefore, the optimal negative pressure value is checked several times during cardiopulmonary bypass. The optimal negative pressure is defined as the minimal value necessary to achieve the higher blood level in the cardiotomy reservoir during full-flow support.

We have noticed that, during cardiopulmonary bypass, a slight increase in the negative pressure often improves venous drainage and prevents the need of additional volume in the reservoir if the blood level in the cardiotomy reservoir reaches the minimal value.

We have published our results on blood transfusion with VAVD and miniaturized bypass circuit [8]. We have chosen to maintain a hemoglobin level of at least 8 g/dl (equivalent to a hematocrit of about 25 %). To estimate the tolerance of this nadir hemoglobin level, we used the perioperative value of blood lactate level and the time to extubation. Peri and post-operative lactate levels were lower in bloodless surgery patients than in transfused patients and the time to extubation was shorter in bloodless surgery patients than in transfused patients. Transfusion should not be considered as the only outcome endpoint. We have sought to do prevent modifications of patient care that are likely to induce new risks and thereby counterbalance the benefit of bloodless surgery.

In our study of 150 procedures, all patients weighing less than 6.4 kg were transfused, with a mean hemoglobin nadir level of 8.7 g/dl while 45 % of the patients weighing between 6.4 kg and 10 kg were transfused. It is noteworthy that:

- These percentages include not only the intra-operative period but also the post-operative period.
- We never used platelet infusions
- Patients were exposed to two different donors in 98 % of cases (i.e. 1 packet of red blood cells and 1 of fresh frozen plasma), one patient was exposed to one donor and one patient to three donors.


Many efforts are needed to decrease the use of allogeneic blood transfusion during pediatric cardiac surgery. We must keep in mind that a volume of 10 ml for a 3 kg baby is equivalent to about 1/25 of his total blood volume and is thus similar to a volume of 200 ml for a 70 kg adult.

VAVD is an efficient technique of assisted venous drainage. It is a low-cost technique without the need for expensive material. Furthermore, the reduction in blood transfusion expected following its use is also likely to cut the cost of pediatric cardiac surgery.

There are a few risks associated with the use of VAVD, which can be prevented. A pressure sensitive security valve must be present on the cardiotomy reservoir. We also recommend a continuous monitoring of the pressure in the cardiotomy reservoir.

It is essential to disconnect the cardiotomy from the vacuum source before weaning from bypass, particularly when using a non-occlusive pump.

To prevent arterial microemboli, air in the venous line should be avoided. Especially during the initiation of bypass, the venous line must be air-free [22]. During CPB, great attention must be paid during manipulation of the oxygenator's sampling manifold, for blood exams or drug injection, to avoid the introduction of air. A gentle injection in the manifold can prevent detection of gaseous microemboli in the arterial line during injection [23-24].

CONCLUSION

VAVD is a simple, efficient and safe technique providing that the specific rules of the technique are well known, well understood and well respected.