Technical considerations in the use of the membrane oxygenator (Lande'-Edwards) for open heart surgery

Experience has shown that successful cardio-pulmonary bypass depends upon an efficient oxygenator which causes minimum blood damage. Experience with 1,600 Lande'-Edwards membrane oxygenators in more than 1,000 patients has led to the development of a simple and safe routine for their use in open heart surgery. The circuitry, method of preparation for perfusion, and management of perfusion are described in detail.     

Development of a membrane oxygenator for ECMO using a novel fine silicone hollow fiber

One of the limitations of conventional silicone hollow fiber oxygenators compared with microporous membrane oxygenators is poor gas permeability. However, the silicone hollow fiber is free from plasma leakage, which is the major life limiting factor of the microporous membrane oxygenator. It has been difficult to fabricate a fine, thin hollow fiber for reduction of resistance to gas permeability because of the poor mechanical strength of conventional silicone materials. The authors developed a novel silicone material with sufficient mechanical strength, and a fine silicone hollow fiber with a diameter of 30 microns and wall thickness of 50 microns, which is approximately half that of a conventional silicone hollow fiber. Using this newly developed silicone hollow fiber, the authors developed a compact extracapillary flow membrane oxygenator. The oxygenator consists of fine silicone hollow fibers inserted in a housing made of polycarbonate. The housing is a cylindrical case, 20 cm long and 55 mm in inside diameter. The hollow fibers are cross-wound. The surface area of the membrane is 2.0 m2, and priming volume is 230 ml. Gas transfer performance of the newly developed oxygenator was evaluated by in vitro experiments. Oxygen and carbon dioxide transfer rates were 195 ml/min and 165 ml/min, at a blood flow rate 3 L/min. The novel silicone membrane oxygenator developed in this study can be used for extended duration in such applications as extracorporeal membrane oxygenation.     

Use of in vitro models of haemofiltration and haemodiafiltration to estimate dosage regimens for critically ill patients prescribed cefpirome

Membrane oxygenators have now gained wide acceptance. A new hollow-fibre membrane oxygenator, the Dideco D903 Avant 1.7, with an optimized membrane surface (1.7 m2) and a wavy blood flow pattern, was tested for gas transfer and blood path resistance in a standardized setting with surviving animals. Three calves (mean body weight 63.29 +/- 2.9 kg) were connected to cardiopulmonary bypass by jugular venous and carotid arterial cannulation, classic roller pump and the Dideco D903 oxygenator with a mean flow rate of 53 +/- 0.1 ml/kg/min for 6 h. After this time, the animals were weaned from the CPB and thereafter from the ventilator. After 7 days, the animals were killed electively. Blood gas analysis was performed before bypass, after mixing (10 min) and then hourly for the 6 h of perfusion. Further samples were taken 30 min (spontaneous breathing) and 60 min after bypass (extubated). Physiological blood gas values could be maintained throughout perfusion in all animals. Mean arterial oxygen saturation varied between 99.3% and 99.7% for the arterial side of the oxygenator compared to 64.6% and 71% for the venous side. The highest mean pressure drop through the oxygenator was 54 mmHg. Postbypass blood gas analysis showed physiological values and no evidence of major lung trauma or pulmonary oedema in relation to the 6 h perfusion. The hollow-fibre membrane oxygenator, Dideco D903, offers excellent gas exchange capabilities and a low pressure drop under experimental conditions, despite reduced membrane surface area. The post mortem examination did not show any deleterious lesion.     

A baboon model for hematologic studies of cardiopulmonary bypass

Objective investigation of new inhibitors of blood protein or cellular systems that are activated during cardiopulmonary bypass (CPB) is impeded by the absence of a satisfactory animal model. Because most baboon hematologic proteins immunologically cross-react with those used for human assays, we developed a robust, reusable baboon model of CPB. Blood samples were obtained from adult baboons at six time intervals before, during, and after 60 minutes of partial CPB at 37 degrees C with peripheral cannulas. Both membrane (n = 7) and bubble oxygenators (n = 7) were investigated. We measured platelet and white blood cell counts; platelet response to adenosine diphosphate and release of beta-thromboglobulin; fibrinopeptide A, prothrombin fragment F1.2, thrombin-antithrombin complex, D-dimer, and plasmin-antiplasmin complex; activated complement (C3b/c and C4b/c); elastase-alpha1 proteinase inhibitor complex; and bleeding times. Adherent glycoprotein IIIa antigen in Triton X-100 washes of the perfusion circuit was also measured. Markers of baboon platelet, complement, and neutrophil activation and thrombosis significantly increased during CPB with bubble oxygenator systems but did not change appreciably in membrane oxygenator circuits. Markers of fibrinolysis, D-dimer, and plasmin-antiplasmin complex did not change with either oxygenator. The baboon model of CPB, when a bubble oxygenator is used, is a robust, reusable animal model for evaluating inhibitors of platelet, complement, and neutrophil activation and thrombosis during and after CPB.     

Effects of blood flow pulse frequency on mass transfer efficiency of a commercial hollow fibre oxygenator

The clinical advantages achievable through pulsatile blood perfusion during cardio-pulmonary bypass have recently suggested the design of new pulsatile systems for extracorporeal circulation. Still it is not clear whether current commercial membrane oxygenators could be adopted with such systems, since their behaviour with pulsatile perfusion has not been satisfactorily documented yet. In a previous paper, we assessed that pulsatile perfusion of a widely used hollow fibre oxygenator (Sorin Monolyth) at 60 bpm provides more time-consistent oxygen transfer than steady perfusion. The present work is aimed to evaluate how the pulse frequency influences the gas transfer performance of the same device. The oxygenator was subjected to in vitro trials using a roller pump with pulsatile module (Stockert Instrumente) to generate pulsed flow. Four different pulse frequencies (45, 60, 75 and 90 bpm) were investigated at a fixed blood flow rate (4.0 l/min). The experiments lasting six hours were carried out using bovine blood with inlet conditions according to AAMI standards requirements. Blood samples were withdrawn every hour and O2 and CO2 transfer rates were evaluated. The experimental findings confirm that with pulsatile blood flow no time decay take place during prolonged perfusion. Moreover, when pulse frequency increases, transition levels occur for both O2 and CO2. Over these thresholds gas transfer rates display significant increases (p < 0.05), though of little magnitude (up to 2.5% for oxygen over 60 bpm; up to 3.7% for carbon dioxide over 75 bpm).     

Development of a silicone hollow fiber membrane oxygenator for ECMO application

A new silicone hollow fiber membrane oxygenator for extracorporeal membrane oxygenation (ECMO) was developed using an ultrathin silicone hollow fiber, with a 300 microm outer diameter and a wall thickness of 50 microm. The hollow fibers were mechanically cross-wound on the flow distributor to achieve equal distribution of blood flow without changing the fiber shape. The housing, made of silicone coated acryl, was 236 mm long with an inner diameter of 60 mm. The surface area was 1.0 m2 for prototype 211, and 1.1 m2 for prototype 209. The silicone fiber length was 150 mm, and the silicone membrane packing density was 43% for prototype 211 and 36% for prototype 209. Prototype 211 has a priming volume of 208 ml, and prototype 209 has a priming volume of 228 ml. The prototype 211 oxygenator demonstrates a gas transfer rate of 120 +/- 5 ml/min (mean +/- SD) for O2 and 67 +/- 12 ml/min for CO2 under 2 L of blood flow and 4 L of O2 gas flow. Prototype 209 produced the same values. The blood side pressure drop was low compared with the silicone sheet oxygenator (Avecor, 1500ECMO). These results showed that this new oxygenator for ECMO had efficiency similar to the silicone sheet oxygenator that has a 50% larger surface area. These results suggest that the new generation oxygenator using an ultrathin silicone hollow fiber possesses sufficient gas transfer performance for long-term extracorporeal lung support.     

Reusable tubular membrane oxygenator for isolated organ hemoperfusion

A reusable tubular membrane oxygenator is described for hypotraumatic hemoperfusion of isolated organs in physiological studies. The constructed oxygenator was of approximately 0.24-m2 effective surface area and contained 450 silicone rubber capillaries of 0.51-mm nominal ID, 34.9 cm long, fixed by conical-shaped, plastic blood headers at manifolds made from Dow-Corning MDX-4-4210 silicone elastomer. During ex vivo hemoperfusions in dogs at inlet hemoglobin saturations near 67%, oxygen transfer rates of the oxygenator increased serially, from 16.6 +/- 1.7 ml/min per m2 (mean +/- SD) at blood flows of 100 ml/min to 34.1 +/- 3.8 ml/min per m2 at flows of 500 ml/min. The oxygenator was thromboresistant and of much loss priming blood volume and wall compliance than the nonresuable Travenol membrane oxygenator of 0.26-m2 effective surface area. The tubular oxygenator was easily cleaned and reassembled, with reproducible oxygen transfer rates. It should prove useful for hemoperfusion studies in organs of moderate size, such as the isolated canine kidney, stomach, and pancreas.     

Effect of alpha(alpha)-cross-linked hemoglobin and pyridoxalated hemoglobin polyoxyethylene conjugate solutions on gastrointestinal regional perfusion in hemorrhagic shock

BACKGROUND: Hemoglobin-based blood substitutes may cause vasoconstriction, which could limit organ perfusion during trauma resuscitation. We investigated the effect of two hemoglobin solutions on regional blood flow and mucosal perfusion in the gastrointestinal tract in a hemorrhagic shock model. METHODS: Twenty-four swine were bled 30% of blood volume over 1 hour. Six additional animals were anesthetized and monitored but did not undergo hemorrhage. Bled animals were resuscitated with alpha(alpha)-hemoglobin (alpha(alpha)Hb), pyridoxalated hemoglobin polyoxyethylene conjugate (PHP), shed blood, or lactated Ringer's solution. Regional blood flow was measured by radiolabeled microspheres. Gastric mucosal perfusion was estimated by measuring intramucosal pH (pHi) by tonometry. RESULTS: PHP and shed blood restored small-bowel flows to sham values, whereas lactated Ringer's solution and alpha(alpha)Hb did not. Shed blood and PHP, but not alpha(alpha)Hb, restored cardiac index (CI) to baseline (p < 0.05). Mean pulmonary artery pressure was elevated over baseline with alpha(alpha)Hb and PHP and remained elevated with alpha(alpha)Hb (p < 0.05). pHi was significantly lower after resuscitation with PHP than with other fluids. CONCLUSION: PHP was efficacious in restoring CI and small-bowel flow, but the pHi remained low, indicating possible continued mucosal ischemia. Alpha(alpha)Hb led to limited recovery of CI and small-bowel blood flow but restored pHi close to baseline. Shed blood was efficacious in restoration of pHi, gastrointestinal blood flows, and systemic hemodynamics.     

Oxygenation in the rabbit myocardium: assessment with susceptibility-dependent MR imaging

PURPOSE: To determine the feasibility of using hemoglobin (Hb) desaturation as an indicator of myocardial oxygenation. MATERIALS AND METHODS: High-resolution gradient-echo nuclear magnetic resonance (MR) images of isolated, blood-perfused rabbit hearts were obtained at various blood oxygenation levels. The hearts were perfused at 37 degrees C with a Langendorff apparatus modified for nuclear MR imaging. The perfusate contained bovine red blood cells in a cardioplegic solution that eliminated motion artifacts and minimized arteriovenous oxygenation differences. Hb saturation was varied (7%-100%) randomly. Perfusion pressure was continuously monitored, and blood samples were obtained. RESULTS: There was a substantial correlation between image signal intensity in the myocardium and Hb saturation in the blood, believed to be due to susceptibility effects of the paramagnetic species deoxyhemoglobin. CONCLUSION: Direct and noninvasive determination of regional Hb saturation with susceptibility-dependent MR imaging may provide information regarding regional myocardial O2 content.     

Lipid peroxidation injury to red blood cells during extracorporeal circulation: mechanism and protection

Twelve dogs were subjected to cardiopulmonary bypass with membrane oxygenator for 120 minutes. The effect of lipid peroxide injury on red blood cells was studied by measurement of plasma and erythrocyte membrane lipid peroxide, deformabioity of erythrocyte, plasma free hemoglobin, superoxide dismutase, Na(+)-K(+)-ATPase and Ca(2+)-Mg(2+)-ATPase of erythrocytes. The effect of vitamin E on red blood cells was also investigated. The findings indicated that vitamin E might protect red blood cells from lipid peroxide injury during extracorporeal circulation. The mechanism of damage effect of lipid peroxide and the protective effect of vitamin E on red blood cells were briefly discussed.     

Prevalence and effects of cardiac rehabilitation on depression in the elderly with coronary heart disease

Stellate ganglion block (SGB) is considered to increase blood flow to the governing area, but recent studies have shown the decrease of common carotid arterial blood flow in the unblocked side following SGB. We investigated the influence of SGB on bilateral cerebral blood flow and oxygenation using a regional oxygen saturation (rSO2) monitor, TOS-96 (TOSTEC, Japan). The changes of blood volume (hemoglobin index: Hb I) and rSO2 were measured in twenty adult patients who underwent SGB at the transverse process of the C 6 vertebra. Prior to the investigation, two sensors were calibrated and placed on either side of each patient's forehead in order to see the difference. There were significant increases in rSO2 and Hb I in the blocked side and decreases in rSO2 and Hb I in the contralateral side. These results suggest that blood flow in the blocked side following SGB increases with decreased blood flow in the unblocked side. Patients with cerebral vascular disease undergoing SGB might be at risk of a decrease in cerebral blood flow in the unblocked area.     

Perfusion and thermal field during hyperthermia. Experimental measurements and modelling in recurrent breast cancer

Recurrences of malignant tumours in the chest wall are proposed as a valuable model of tissue mainly perfused by small size vessels (the so-called 'phase III' vessels). Invasive thermal measurements have been performed on two patients affected by cutaneous metastasis of malignant tumours during hyperthermic sessions. Thermal probes were inserted into catheters implanted into the tissue at different depths. In one of the catheters a probe connected with laser-Doppler equipment was inserted to assess blood perfusion in the tumour periphery. The perfusion was monitored throughout the sessions, and a noticeable temporal variability was observed. The effect of the perfusion on the thermal map in the tissue was evaluated locally and the 'effective conductivity' of the perfused tissue was estimated by means of the numerical integration of the 'bio-heat' equation. The tumour temperature, at the site where the perfusion probe is located, can be predicted by the numerical model provided two free parameters, alpha and beta, are evaluated with a fitting procedure. Alpha is related to the effective conductivity and beta to the SAR term of the bio-heat equation. The model aimed at estimating the 'effective conductivity' K(eff) of the perfused tissue, and average values of K(eff) of 0.27 +/- 0.03 W m(-1) degrees C(-1) in Patient 1 and of 0.665 +/- 0.005 W m(-1) degrees C(-1) in Patient 2 were obtained throughout the treatment. However, when the average temperature in a larger tumour volume is to be predicted but only a single, 'local' measurement of the perfusion is available and is assumed to be representative for the whole region, the model results are far less satisfactory. This is probably due to the fact that changes of blood perfusion throughout hyperthermic sessions occur to different extents within the tumour volume, and the differences in perfusion cannot be ignored. The above result suggests that, in addition to the 'temperature map', also a 'perfusion map' within the heated volume should be monitored routinely throughout hyperthermic sessions.     

Hyperthermic isolated limb perfusion with cis-diamminedichloro-platinum. II. An experimental study in dogs with a balloon-occlusion technique for repeated high-dose treatment

Isolated organ perfusion is attractive for regional high-dose chemotherapy because of its advantage to reduce whole body toxicity. Intraoperative hyperthermic isolated perfusion procedures involving a heart-lung machine have been developed, but repeated treatments carry a high risk of vessel and tissue damage. Therefore, a study of isolated hyperthermic limb perfusion in four dogs was conducted using a balloon-occlusion technique including a hyperthermia unit, two low-flow rotary pumps, a bubble oxygenator, and two polyurethane balloon catheters. After 15 min infusion of cisplatinum the concentrations of serum platinum (Pt) in the isolated limb and in the whole body were measured by atomic absorption spectroscopy. Regional exposure to Pt was more than 10-fold higher than systemic exposure. After 60 min isolated limb perfusion, the area under the curve (AUC) of Pt concentrations in the isolated limb showed values between 767.4 and 1055.6 micrograms/l x 60 min, whereas in the whole body values between 59.8 and 75.9 micrograms/l x 60 min were obtained. Repeated isolated limb perfusions with the balloon-occlusion technique were performed in three dogs without systemic side effects. This model of regional chemotherapy may be useful for preoperative chemotherapy in malignant tumors of the limbs.     

Study about development of extracapillary blood flow type oxygenator--comparison with HF-4000, Capiox II, and Silox]

Results about comparison of new extracapillary blood flow type polypropylene hollow fiber oxygenator (PHO) with HF-4000, Capiox II and Silox for effects on blood components, hemolysis, immunoglobulin and high molecular weight proteins, such as complements: 1) Platelet protection was about equal. Silox was best in leucocyte protection. 2) PHO was superior to HF-4000. The extracapillary type was better for hemolysis. 3) Silox was best in immunoglobulin protection. Effect on C3 and C4 was equal, and Silox was best in protection of CH50. Slight differences were caused by silicone membrane, not caused by blood flow type. 4) PHO was superior to HF-4000 (commercial extracapillary type), and almost equal to Capiox II (intracapillary type).     

MR evaluation of developmental venous anomalies: medullary venous anatomy of venous angiomas

With continuous enhancement in oxygenator design, the question is raised as to how these changes actually impact the performance of the oxygenator. The recent addition of two new oxygenators by the Bentley Division of Baxter Healthcare Corporation provided us with a unique opportunity to compare the performance of each device and isolate the impact of each design change on performance. While the basic design and flow patterns have remained the same, application of the Duraflo II treatment has produced the Univox Gold and a change in the fibre-winding technique has produced the SpiralGold. This study compared the effects of heparin coating (Univox to Univox Gold) and fibre-winding (Univox Gold to SpiralGold) on gas and heat transfer and resistance to blood flow (pressure drop). Six oxygenators of each model were evaluated utilizing an in vitro single pass circuit, which first conditioned bovine blood to the Association for the Advancement of Medical Instrumentation (AAMI) venous standards. Blood flows of 4.0, 5.0, 6.0 and 7.0 l/min, FiO2 values of 1.0, 0.8 and 0.6, and gas-to-blood flow ratios of 0.5, 1.0 and 1.5 were chosen as test variables. Data generated included oxygen transfer, carbon dioxide transfer, arterial pO2, resistance to blood flow, and coefficient of heat exchange. The results indicate that the Duraflo II treatment does not have a significant effect on gas and heat transfer or resistance to blood flow. The fibre-winding technique employed with the new SpiralGold, however, has improved significantly gas exchange and arterial pO2 when compared with the previous Univox models. Resistance to blood flow and coefficient of heat exchange were not affected significantly by the winding technique.     

Spatial and temporal analysis of human motor activity using noninvasive NIR topography

The effect of motor activity on the left fronto-central region of the human brain was analyzed spatially and temporally by using noninvasive near-infrared light (NIR) topography. The changes in oxygenation states caused by motor activity were measured using intensity-modulated NIR spectroscopy at ten measurement positions on the head surface. The subject randomly performed unilateral finger opposition for 30 s as motor stimulation. When the subject performed contralateral (right) finger movement, significant increases in both oxygenated hemoglobin (oxy-Hb) and total hemoglobin (total-Hb) and decreases in deoxygenated hemoglobin (deoxy-Hb) were observed in a particular area. By mapping the static topograms of the changes of each Hb and comparing them with an anatomical image of MRI, it was found that the particular area was located on the motor cortex along the central sulcus. By mapping the dynamic topograms of the changes of total-Hb, which reflect the cerebral blood volume, and analyzing the spatiotemporal hemodynamic changes associated with the brain activity, it was found that the regional change in cerebral blood volume in the primary motor area overlaps the global change around the motor cortex. These results demonstrate that NIR topography can be used to effectively observe the human brain activity.     

Safe levels of hemoglobin concentration in autologous blood transfusion

It has been generally accepted that an adequate oxygen-carrying capacity can be achieved with a hemoglobin concentration of 7 g/dl, as far as the patient's intravascular volume is sufficient to allow tissue perfusion. To guarantee patient's safety in the operating theater, patient's oxygenation, ventilation, circulation and temperature, which enable oxygen utilization in tissues, should be monitored vigilantly and ensured strictly. This is also true when taking care of anemic patients outside the operating theater, because failure of these functions in anemic patients leads directly to tissue hypoxia. Besides the standard monitoring, measuring oxygen carrying capacity/consumption parameters and gastric/sigmoidal intramucosal pH have been shown to be helpful to estimate tissue oxygenation. Therefore, safe levels of hemoglobin concentration should be determined according to the ability of doctors and nursing staffs to evaluate and to maintain patient's systemic and tissue oxygenation as well as to the patient's pathophysiological conditions.     

Evaluation of an extracorporeal membrane oxygenation system using a nonporous membrane oxygenator and a new method for heparin coating

A new heparin binding method was applied to a miniature extracorporeal membrane oxygenation (ECMO) system with a nonporous membrane oxygenator (the priming volume, 45 mL; the membrane surface area, 0.4 m2; maximal flow rate, 2 L/min) that is resistant to plasma leakage. The authors evaluated the stability of the immobilized heparin in vitro and the feasibility of this system in animals. Samples of hollow fibers and tubing were rinsed at 40 degrees C for 4 days in normal saline, Ringer's lactate, and 1 mol/L NaCl solution. Heparin activities on hollow fibers after rinsing were 99 +/- 2.3% (mean +/- SD), 96 +/- 3.9%, and 93 +/- 2.0% of the control in each solution, while those of the tubing were 87 +/- 4.1%, 86 +/- 3.1%, and 76 +/- 8.6%, respectively. Veno-arterial ECMO using this heparin-coated system was performed on five beagles (8 to 12 kg) for 10 hours. Neither major thrombus formation nor plasma leakage was detected during the procedure in spite of a low flow rate (300 mL/min) and a reduced activated clotting time (mean, 128 seconds). Platelets decreased to 52% of the control (P < .01) at 1 hour, but no progressive decrease was seen thereafter. Antithrombin-III decreased (P < .01) and thrombin/antithrombin III complex increased (P < .05 at 4 hours and P < .01 at 6, 8, and 10 hours) during bypass, but the changes of fibrinogen and fibrinopeptide A were not significant. Fibrinogen/fibrin degeneration products, fibrinopeptide B beta 15-42, and plasma-free hemoglobin levels did not rise significantly. O2 transfer of the oxygenators at a flow rate of 300 mL/min were 12.3 +/- 0.4 mL/min at 30 minutes, 14.3 +/- 1.2 mL/min at 5 hours, and 14.7 +/- 1.7 mL/min at 10 hours (no statistical difference). Histological examination of the brains and the kidneys showed no evidence of thromboembolic sequela in any of the animals. These results suggest that this new system is a promising device for long-term ECMO in which the amount of systemic heparinization can be reduced with the minimal possibility of plasma leakage.     

Activation status and function of the VLA-4 (alpha4beta1) integrin expressed on human melanoma cell lines

A highly sensitive microspectrophotometer was developed to measure spectral changes of oxyhemoglobin (oxy Hb) in single red blood cells (RBC) incubated with stimulated macrophages as a model of nitric oxide (NO) dependent cytotoxicity. Our microspectrophotometer, using a modified acousto-optic tunable filter (AOTF) and a 2-dimensional CCD array, allows fast spectrophotometric data acquisition. Human RBC treated with various concentrations of NO showed spectral changes due to the conversion of oxy Hb to methemoglobin (met Hb), in which the change in absorption differences at alpha (557 590 nm) and beta (542-525 nm) bands showed a linear relationship with the concentration of NO up to 100 microM. In contrast to highly diffusible NO, nitrite ions (NO2-) seem to enter RBC very slowly, resulting in negligible formation of met Hb in the presence of 5 mM glucose even during a prolonged incubation period. RBC were incubated with murine macrophages with and without lipopolysaccharide (LPS) in the presence of glucose for 24 and 40 h and subjected to the microspectrophotometric assay. The RBC incubated with LPS-stimulated macrophages showed significant changes in the spectrum due to NO-dependent conversion of oxy Hb to met Hb, which corresponded to the spectral changes of RBC treated with a several times higher concentration of NO than that in the culture medium. The trapping efficiency was calculated from the amounts of the NO released from macrophages and of the met Hb formed in the RBC, which gave a high efficiency (43%). The results suggest that RBC trap NO directly by cell cell interaction with macrophages. This spectrophotometric system is available for use with just a few drops of samples to study NO-specific cytotoxicity as a model of RBC without the use of any chemical reagent, in parallel with microscopic observations on changes of the cellular morphology under physiological conditions, such as membrane damage leading to hemolysis, adherence, and phagocytosis.     

Effects of an intravenous injection of NPY on leptin and NPY-Y1 receptor mRNA expression in ovine adipose tissue

We have previously shown significant bubble formation in Medtronic Maxima hard-shell venous reservoirs (HSVRs). In the present study, we not only investigated the mechanism of this bubble formation, but also the extent of bubble clearance by membrane oxygenators and arterial line filters. In addition, we also compared the performance of five HSVRs with respect to bubble formation and venous air filtration. Salvaged clinical CPB circuits containing different HSVRs were studied by downstream Doppler monitoring under fixed flow-decreasing volume, fixed volume-increasing flow, and entrained venous air conditions. Bubbles formed in the Medtronic Maxima top entry HSVR at volumes below 800 ml and flows above 3.5 l min-1, and were incompletely removed by a membrane oxygenator and arterial line filter. Decreased bubbling was seen when the reservoir atmosphere was flushed with CO2, suggesting that these bubbles formed in a fountain at the venous inflow. The Medtronic Maxima Forte HSVR formed significantly fewer bubbles at low volumes, and filtered venous air effectively. Negligible bubble formation occurred in the Sorin, Terumo, or Baxter reservoirs. The minimum recommended operating volume for the Medtronic Maxima top entry reservoir should be reset at 600 ml and this device should always be used with an arterial filter. Bubble formation is substantially reduced in the new Medtronic Maxima Forte HSVR and this device is a good filter for venous air.