Scott I. Merz

Nitric oxide releasing silicone rubbers with improved blood compatibility: preparation, characterization, and in vivo evaluation.

Nitric oxide (NO) releasing silicone rubbers (SR) are prepared via a three-step reaction scheme. A diamino triaminoalkyltrimethoxysilane crosslinker is used to vulcanize hydroxyl terminated polydimethylsiloxane (PDMS) in the presence of ambient moisture and a dibutyltin dilaurate catalyst so that the respective diamine triamine groups are covalently linked to the cured SR structure. These amine sites are then diazeniumdiolated, in situ, when the cured SR is reacted with NO at elevated pressure (80 psi). Although nitrite species are also formed during the NO addition reaction, in most cases the diazeniumdiolated polymer is the major product within the final SR matrix. Temperature appears to be the major driving force for the dissociation of the attached diazeniumdiolate moieties, whereas the presence of bulk water bathing the SR materials has only minimal effect on the observed NO release rate owing to the low water uptake of the SR matrices. The resulting SR films/coatings release NO at ambient or physiological temperature for up to 20 d with average fluxes of at least 4 x 10(10) mol x cm(-2) x min(-1) (coating thickness > or = 600 microm) over first 4 h, comparable to the NO fluxes observed from stimulated human endothelial cells. The NO loading and concomitant NO release flux of the SR material are readily adjustable by altering the diamine triamine loading and film/coating thickness. The new NO releasing SR materials are shown to exhibit improved thromboresistance in vivo, as demonstrated via reduced platelet activation on the surface of these polymers when used to coat the inner walls of SR tubings employed for extracorporeal circulation in a rabbit model.

Reduced platelet activation and thrombosis in extracorporeal circuits coated with nitric oxide release polymers

Objective: To determine whether the use of nitric oxide (NO)‐releasing polymers coated onto the inner surface of extracorporeal circuits can reduce platelet consumption and activation in the absence of systemic heparinization using a rabbit model of venovenous extracorporeal circulation. Design: Prospective, controlled trial Setting: Research laboratory at an academic medical institution. Subjects: New Zealand White Rabbits Interventions: Anesthetized, tracheotomized, and ventilated New Zealand White rabbits were injected with freshly prepared, 111In(oxine)3 labeled single donor platelets through the external jugular vein. After baseline measurements, these animals were placed on venovenous extracorporeal circulation through a 1‐m control circuit or NO test circuit for 4 hrs at a blood flow rate of 109‐118 mL/min via roller pump. Four groups were studied: systemically heparinized control circuits, systemically heparinized NO test circuits, nonheparinized control circuits, and nonheparinized NO test circuits. Platelet counts, fibrinogen levels, and plasma free indium levels were measured hourly. Circuits were rinsed and retained for gamma counting after the 4‐hr run or when the circuit clotted. Four animals, one from each group, did not receive radiolabeled platelets so that the circuits could be preserved for scanning electron microscopic examination after the 4‐hr study. Measurements and Main Results: Platelet consumption was significantly reduced in both the heparinized and nonheparinized NO test groups when compared with the controls (p < .0001 and p < .0004, respectively). Platelet adhesion to the extracorporeal circuits was significantly reduced in the nonheparinized test circuits when compared with the controls (p < .05). Scanning electron microscopic examination of the circuits revealed that in the absence of heparin and in the presence of a NO‐releasing surface, platelets retained their spherical nonactivated shape. Conclusions: The incorporation of NO into the surface of extracorporeal circuits reduces platelet consumption and eliminates the need for systemic heparinization in a rabbit model of extracorporeal circulation.

Development and application of a simplified liquid ventilator

OBJECTIVE Perfluorocarbon liquid ventilation has been shown to have advantages over conventional gas ventilation in premature newborn and lung-injured animals. To simplify the process of liquid ventilation, we adapted an extra-corporeal life-support circuit as a time-cycled, volume-limited liquid ventilator. DESIGN Laboratory study that involved sequential application of gas and liquid ventilation in normal cats and in lung-injured sheep. SETTING A research laboratory at a university medical center. SUBJECTS Eight normal cats weighing 2.7 to 3.8 kg (mean 3.1 +/- 0.5), and four lung-injured young sheep weighing 10.4 to 22.5 kg (mean 15.9 +/- 5.0). INTERVENTIONS Normal cats were supported with traditional gas ventilation for 1 hr (respiratory rate 20 breaths/min, peak inspiratory pressure 12 cm H2O, positive end-expiratory pressure 4 cm H2O, and FIO2 1.0). The lungs were then filled with perfluorocarbon (30 mL/kg) and tidal volume liquid ventilation was instituted, utilizing a newly developed liquid ventilation device. Liquid ventilatory settings were 4 secs for inspiration time, 8 secs for expiration time, 5 breaths/min for respiratory rate, and 15 to 20 mL/kg for tidal volume. Liquid ventilation utilizing this device was also applied to sheep after induction of severe lung injury by right atrial injection of 0.07 mL/kg of oleic acid, followed by saline pulmonary lavage. Extracorporeal life support was instituted to provide a stable model of lung injury. For the first 30 mins of extracorporeal support, all animals were ventilated with gas. Animals were then ventilated with 15 mL/kg of perfluorocarbon over the ensuing 2.5 hrs. MEASUREMENTS AND MAIN RESULTS In normal cats, mean PaO2 values after 1 hr of liquid or gas ventilation were 275 +/- 90 (SD) torr (36.7 +/- 10.4 kPa) in the liquid-ventilated animals and 332 +/- 78 torr (44.3 +/- 10.4 kPa) in the gas-ventilated animals (NS). Mean PaCO2 values were 40.5 +/- 5.7 torr (5.39 +/- 0.31 kPa) in the liquid-ventilated animals and 37.6 +/- 2.3 torr (5.01 +/- 0.31 kPa) in the gas-ventilated animals (NS). Mean arterial pH values were 7.35 +/- 0.07 in the liquid-ventilated animals and 7.34 +/- 0.04 in the gas-ventilated animals (NS). No significant changes in heart rate, mean arterial pressure, lung compliance, or right atrial venous oxygen saturation were observed during liquid ventilation when compared with gas ventilation. In the lung-injured sheep, an increase in physiologic shunt from 15 +/- 7% to 66 +/- 9% was observed with induction of lung injury during gas ventilation. Liquid ventilation resulted in a significant reduction in physiologic shunt to 31 +/- 10% (p < .001). In addition, the extracorporeal blood flow rate required to maintain the PaO2 in the 50 to 80 torr (6.7 to 10.7 kPa) range was substantially and significantly (p < .001) lower during liquid ventilation than during gas ventilation (liquid ventilation 15 +/- 5 vs. gas ventilation 87 +/- 15 mL/min/kg). CONCLUSIONS Liquid ventilation can be performed successfully utilizing this simple adaptation of an extracorporeal life-support circuit. This modification to an existing extracorporeal circuit may allow other centers to apply this new investigational method of ventilation in the laboratory or clinical setting.

Plasma leakage through microporous membranes : role of phospholipids

Plasma leakage through microporous membrane oxygenators is a well known complication of prolonged extracorporeal circulation. The authors hypothesized that adsorption of bipolar plasma molecules, such as phospholipids on the microporous membrane, results in formation of a hydrophilic layer over the hydrophobic surface of the membrane; this, in turn, leads to plasma leakage at normal surface tensions. A lipid phosphorus assay was used to measure phospholipid adsorption onto the fibers of microporous membrane oxygenators tested under a variety of experimental conditions. Adsorption of phospholipids on the microporous membrane was concentration dependent. Reproducible plasma leakage occurred both in vitro and in vivo, and the time to leakage was dependent on the concentration of phospholipids adsorbed upon the microporous membrane. Based upon these results, the authors conclude that adsorption of phospholipids contributes to the development of plasma leakage through microporous membrane oxygenators.

Development of an ambulatory artificial lung in an ovine survival model.

We are developing an artificial lung (AL) as an eventual bridge to lung transplant or recovery. The device is rigidly housed, noncompliant, and has a very low resistance to blood flow. In eight sheep, arterial cannulae were anastomosed end-to-side to the proximal and distal main pulmonary artery, and attached to the AL. A pulmonary artery snare between anastomoses diverted full pulmonary blood flow through the AL. Eight of eight sheep survived the preparation. Mean pressure gradient across the AL was 8 mm Hg (3 Wood units; 8 mm Hg/2.8 L/min). Four of eight sheep tolerated immediate full diversion of blood flow and died at 24 and 40 hours (exsanguination) or 168 and 168 hours (elective sacrifice). Four of eight sheep were intolerant of full flow: two died of right heart failure at <8 hours with full flow through the device (full snare); the other two survived with partial device flow (partial snare), but the device clotted. These two then underwent successful closed-chest cannula thrombectomy and device change-out at 53 and 75 hours, and subsequently tolerated full flow. Long-term (up to 7 day) survival with complete diversion of pulmonary blood flow through a noncompliant, low-resistance AL is possible. Initial right heart failure in this model was 50% (4 of 8).

A standardized system for describing flow/pressure relationships in vascular access devices.

Catheters are usually characterized by length and external diameter only. We developed a system to describe flow-pressure characteristics with a single number, M, patterned after a Reynolds number-friction factor correlation. A simple alignment chart was constructed that can be used to determine M for catheters of regular internal diameter and given length. If tapers, side holes, or integral connectors are present, M can be determined in-vitro with water by measuring pressure drop at various flows. The chart describes flow pressure functions for blood (Hct = 0.41) if M is known. Using this system, any catheter, needle, valve seat, extracorporeal tubing, or vascular graft can be assigned a single number that describes hydrodynamic performance characteristics.

Laboratory evaluation of a double lumen catheter for venovenous neonatal ECMO

The authors designed and tested a 14F outside diameter thin-walled double lumen catheter (DLC) for neonatal venovenous (VV) extracorporeal membrane oxygenation (ECMO). In vitro tests with water and dye solution showed capacity of the drainage lumen was 1,096 ml/min at 100 cm siphon, and pressure drop across the perfusion lumen was 300 mmHg at 500 ml/min flow. Recirculation at 500 ml/min flow ranged from 5 to 29%, depending upon simulated cardiac output. The highest serum hemoglobin during 12 hour 400 ml/min flow VV bypass in five dogs was 49 mg/dl. Typical oxygen transport in four dogs was 25 cc/min at 400 ml/min flow. This catheter is well suited for clinical VV ECMO in neonates.

Laboratory experience with a novel, non-occlusive, pressure-regulated peristaltic blood pump

Current blood pumps have potential safety problems, including the ability to generate extreme positive and negative pressures. These problems were addressed in the design and testing of a non-occlusive, peristaltic blood pump. The pump consists of a tubing of unique design (pump chamber) wrapped under tension around a rotor with rollers. The pump chamber design is such that the pump is passively filling; flow is dependent upon the pressure of the blood supply and the size of the pump chamber. Thus, negative pressures cannot be generated. With the outlet occluded, the pump produces the maximum attainable pressure, which can be set by adjusting the tension of the pump chamber around the rollers. The design characteristics make the pump suitable for prolonged use. The pump was tested in vitro for pressure safety, hemolysis, and durability. The pump prototype was used in 25 experiments involving extracorporeal circulation on sheep, with an average duration of approximately 6 hr and bypass flow rates between 0.5 and 2.0 L/min. No pump related complications occurred in any of these experiments. The pump described here is suitable for short- and long-term perfusion applications, and does not require additional pressure regulation, as do current blood pumps.

New method for describing the performance of cardiac surgery cannulas

Cardiac surgery cannulas are characterized by external diameter only, which provides little information about the pressure-flow characteristics of a cannula. A system has been developed to describe pressure-flow characteristics with a single, unitless number, M, which is patterned after a Reynolds friction factor correlation. A cannula with a lower M number has a more favorable pressure-flow relationship. The M number was determined for 16 arterials cannulas ranging in size from 10F to 26F and 27 venous cannulas sized 12F to 36F. Pressure-flow characteristics vary considerably among cannulas from different manufacturers despite having similar French sizes. Clinical decisions regarding choice of cannula can be simplified by using the M number, which gives a more accurate description of the performance characteristics of a cannula than the French size designation.

An in vitro method for assessing biomaterial-associated platelet activation.

The development of a nonthrombogenic artificial surface for use with indwelling sensors or catheters remains an elusive goal despite decades of ongoing research. In vivo studies are both labor intensive and costly, and are therefore an inefficient way to rapidly screen possible surface materials. The following in vitro model used glass, polyvinyl chloride (PVC), and polypropylene test tubes incubated with 111In-labeled rabbit platelets and illustrated that, despite equivalent platelet count and function, platelet adhesion was greatest on glass (n = 13), with PVC (n = 17) at 67 ± 8% and polypropylene (n = 13) at 43 ± 5% when compared with glass. Extrapolating this method by coating test tubes with new, nonthrombogenic materials is a quick and reliable way to screen material before embarking upon more lengthy in vivo animal studies.