The rat sciatic nerve was transected mid-thigh and grafted with a silicone tube, the central 10 mm of which was filled with a collagen-glycosaminoglycan (CG) matrix. The rats were grafted contralaterally with empty silicone tubes as controls. The earliest compound muscle action potentials (CMAP's) at the plantar muscles were recorded around 11 weeks. After 30 weeks, the distal motor latencies recovered to about 50% higher than normal, the conduction velocity to about 50% normal, and the amplitudes of the CMAP's to about 50% normal. Of 7 rats in this study, all 7 nerves grafted with the CG matrix exhibited recovery, while only 1 grafted with the empty tube exhibited recovery. The CG matrix therefore appears to promote functional nerve regeneration across extended distances.

We have developed a quantitative model which allows assessment of the number of primary motor and sensory neurons which regenerate an axon through a tubular prosthesis bridging a 4–5 mm gap of peripheral nerve (1). Labeled somas of motor and sensory neurons are quantified following retrograde transport of horseradish peroxidase (HRP) applied to the distal stump beyond the bridged transection site. The number of myelinated axons at mid-guide level is also quantified. We have studied biodegradable and nin-biodegradable materials as well as the effects on nerve regeneration of basement membrane materials, and specific proteins added to the nerve guide., Adding a laminin-oontaining gel (Matrigel, Collaborative Research, MA) to the lumen of polyethylene tubes increases the rate of nerve regeneration as well as the maximum distance that can be successfully bridged. Similar effects are seen with collagen-based nerve guides when the laminin-cxitaining gel is incorporated into the tube wall. Addition of highly purified acidic fibroblast growth factor (aFGF) to the lumen of polyethylene tubes filled with collagen significantly increases the number of myelinated axons at mid-tube level and the number of primary sensory neurons labeled with HRP.

We coxpared collagen-based nerve guides of different wall porosities. Tubes freely permeable to bovine serum albumin (BSA) contain significantly more myelinated axons (4 week survival) than identical collagen tubes permeable to glucose-size molecules. These and additional experiments demonstrate that manipulation of the extracellular microenvironment of regenerating PNS axons can significantly enhance the early regenerative response.

Electrodes for FNS were fabricated by photolithographic techniques from foils of Mo, Ta, and stainless steel with lateral dimensions of 1–3 mm and thickness of about 50 μm. More recent designs have utilized a Si carrier with either a Pt or diffused Si track terminated by Au electrodes. Biocompatibility studies conducted on frogs and rabbits indicate good stability of electrodes coated with SiO by plasma deposition. Carriers did not move after implantation and did not show signs of corrosion. The diffused Si design is now preferred since electrode tracks are sub-surface, tracks may be electrically and biologically insulated by SiO2, all fabrication steps are compatible with standard integrated circuit processing such that electronics may be integrated on the carrier with the electrode tracks, and small dimensions are easily attained.

The charge injection limits of a variety of thin film redox materials considered for use as neural stimulation electrodes were evaluated under standardized conditions. Materials tested included the oxides of Ni, Ir, Rh, Ru, and Mn as well as conductive polypyrrole films. Electrodes with geometric areas of, ∼10–4 cm2 were tested in bicarbonate buffered saline of pH 7.3. An electrochemically “safe” potential window was determined for each material by cyclic voltammetry. Charge injection capabilities within that window are evaluated with constant current 0.2 msec pulses. Pulse modes include anodic first and cathodic first biphasic pulses and monophasic cathodal pulses applied to electrodes held at an anodic bias. Pulsing from the anodic bias increased the charge limits for all the materials, but the most dramatic improvement was for Rh oxide. The highest charge injection values were obtained with oxides of iridium and rhodium and were of comparable value for the anodic bias mode of pulsing. Where the effect of film thickness was examined, the charge injection limit leveled off at 2–4 mC/cm2 geometric as the film thickness increased.

Clinical trials with the UCSF/Storz multichannel cochlear implant began in February 1985. The final patient in this series was implanted in June 1986. The objective of the clinical trial was to evaluate the strategies developed by the UCSF implant team through neurophysiological, histopathological, and psychophysical investigations over the past seventeen years. A total of 16 patients were implanted with this four channel unit, a vocoder-based, compressed-analog system. Results are measured with a battery of speech reception tests, including routine audiometric tasks used to evaluate standard clinical populations. Assessments are administered using conventional audiologic procedures. Tests include monosyllabic word recognition (the standard clinical “speech discrimination” list) and the recognition of key words in sentences. Eleven of the 16 patients have now had their one-year postoperative evaluations, permitting observations on their performance over time. Of these 11 patients, ten (91%) are able to obtain a considerable degree of auditory speech recognition. Relative to other cochlear implant projects, this represents an extremely high proportion of the implant population achieving some level of auditory-only speech understanding. The UCSF/Storz clinical trial has helped to clarify and define the specifications for our next generation implant, an elecrically transparent sixteen channel system.

One of the most crucial requirements since the inception of the artificial heart program has been the need to develop biomaterials specific to chronic implantation. Major problems include the propensity for thrombus formation at the blood-materials interface, long-term flexing reliability, and the tendency of the body to reject foreign objects.

Microcellular polymer foams are of interest as biomaterials for uses such as artificial skin, synthetic vascular replacements, artificial organs, and for use in permeable drug delivery systems. We wish to report the use of thermally - induced phase separation as a technology by which a variety of polymer foams can be prepared. This process permits control of foam morphology, cell size, and density. Importantly, the phase separation process results in fully open-celled foam structures. Additives such as small molecules and functionalized polymers have been incorporated into these microcellular foams resulting in polymer surfaces with controlled porosity and chemical modification. Microcellular polyurethane foams, with chemically modified surfaces, have potential application as small-diameter vascular replacements.

Sorption processes (adsorption, absorption, permeation) are of considerable importance in the physico-chemical and biological performance of polymeric biomaterials, especially in cardiovascular applications, and in various controlled drug release and drug carrier systems [1,2]. Transport of molecules of widely ranging molecular weights through synthetic as well as biologic membranes represents a basic process in the performance of many medical devices.

The influence of mechanical properties on the acute and chronic success of vascular grafts remains the subject of considerable debate. It is widely held that the matching of some set of mechanical properties at the anastomosis between a natural vessel and a vascular graft will optimize the likelihood of clinical sucess. Investigations in this area have largely concentrated on compliance (defined as the percentage change in diameter per unit change in pressure) as the parameter to be matched across the anastomosis. However, compliance represents only one out of a set of candidate parameters (including elastic moduli, Poisson ratios, admittance, distensibility, and hydraulic impedance among others) and there is no definitive evidence that it is in any way the most important mechanical parameter to be considered. As well, the concentration on compliance has resulted in the failure to examine the importance of the longitudinal behavior of the vascular grafts. The longitudinal properties of any vessel can profoundly influence the observed circumferential behavior; however, this is, at best, only implicit in the onedimensional compliance parameter.

Coagulase-negative staphylococci are major pathogens in prosthetic implant infections, and the capability of certain strains to produce glycocalyx enhances their infectivity. We investigated the importance of glycocalyx in an animal model of prosthetic vascular graft infection. Coagulasenegative staphylococci (5×108) and 1 cm2 pieces of Dacron or polytetrafluoroethylene (PTFE) grafts were placed in subcutaneous pockets of adult Sprague-Dawley rats. Two strains of bacteria were used, one producing and the other not producing glycocalyx. At 5 days, uninfected grafts were incorporated into the surrounding tissues whereas infected grafts were surrounded by fluid. Grafts were excised and adherent bacteria dislodged by sonication and quantitated. Scanning electron microscopy confirmed the presence of bacteria and glycocalyx, and the efficiency of bacterial dislodgement by sonication. The bacterial adherence (CFU/cm2 mean ± SE) of glycocalyx-producing staphylococci was 3.7±1.1×107 and 3.5±1.0×105 to Dacron and PTFE, respectively (both < 0.05). Nonglycocalyx producing bacteria had decreased adherence to both Dacron (1.4±0.8×104) and PTFE (0.9±0.4±104). Antibiotic administration reduced the adherence of the nonglycocalyx producing strain by one log to both materials (p<0.05) but did not affect the glycocalyx-producing strain. These data demonstrate that glycocalyx production enhances the adherence of coagulase-negative staphylococci to prosthetic vascular graft materials. Prophylactic antibiotic administration does little to inhibit the adherence of glycocalyxproducing bacteria but does affect nonglycocalyx producing strains.

A rapid accurate method has been developed utilizing fiber optic scanning densitometry to quantify the radiopacity of vascular catheters. The technique provides for computerized calculation of relative catheter radiopacity and an appropriate control standard. A densitometer with a 180 degree collection angle for diffuse transmission density measurements was selected based on the diffusing nature of x-ray film (Q-factor 1.80). A benchmark catheter and 2 mil thick brass shim stock were selected as control standards for evaluation of mono- and multilumen tubing using standard x-ray conditions and an aluminum block attenuator.

Results from reproducibility studies show scan-to-scan repeatability is within ±1%, and day-to-day variability is less than 5%. Application studies demonstrated a linear relationship between percent barium sulfate loading and the radiopaqueness of 16 gauge monolumen tubing. Results were also obtained from a clinical chest x-ray film showing good in vivo/in vitro correlation. The amount of radiopaque filler, design and orientation of the radiopaque region of the catheter was found to be critical in obtaining optimum catheter radiopacity.

This method provides a quantitative technique which permits precise evaluation of vascular access devices as compared to the qualitative evaluation provided by methods such as ASTM F 640.

There have been in excess of 90 implantations of several different types of artificial heart (TAH) into patients. The bulk of these implantations have been using the Jarvik TAH as a bridge to provide interim circulatory support for the failed heart while a donor heart is being located and transported. Attempt to use the Jarvik TAH as a permanent, long-term heart substitute were complicated by hemorrhaging, sepsis, peripheral-organ failure and strokes [1]. These latter sequelae are thought to be due to thromboembolism and this concept is supported by the examination of explanted TAH from Barney Clark, the first recipient of the Jarvik TAH, which revealed large thrombus formation around the valve seats and microthrombi on the diaphragm that may have been the sites of thromboemboli [2]. In the case of bridges, the implantation periods are shorter, averaging 10 to 14 days but still a significant number of these patients have had thromboembolic events and explanted TAH have had microscopic aggregations of platelets and thrombi [3,4]. The response of blood to the constituent biomaterials of all cardiovascular devices is highly variable [5]. The documentation of thrombus on a biomaterial surface does not prove it will detach nor does the lack of thrombus on a surface indicate that the surface is benign, as thrombus could have formed and embolized. We have used two approaches to study the dynamic process of thrombogenesis and thromboembolism, in vivo and in vitro. Whole arterial blood samples were analyzed for thromboemboli as an in vivo assessment and epifluorescent videomicroscopy was used to monitor the mechanism of thrombosis, in vitro. Microparticles were found in the arterial circulation of patients with a Jarvik TAB and platelet thrombi were observed to form and embolize from Biomer-coated surfaces during perfusion.

Sulfonated polyetherurethanes, synthesized by the substitution of 0, 5, 10, 15, and 20% of the urethane nitrogens with propane sultone, were evaluated in this study. The water absorption properties are dramatically affected by the sulfonate content. The surface properties are also found to be influenced by the percentage of sulfonate incorporation. The blood-contacting properties, as determined by both an acute and chronic canine ex vivo experiment, show increased thromboresistance with increased propyl sulfonate incorporation. A unique characteristic of the highly sulfonated polymers is the negligible platelet activation and spreading observed using scanning electron microscopy.

The primary role of a red cell substitute is to deliver oxygen to cells either in vivo or in vitro. It seems reasonable to mimic evolution, which solved the problem of oxygen delivery in many species by encapsulating oxygen carrying proteins in cell-sized delivery systems. We have successfully synthesized and tested an artificial red cell (Neohemocytes: see Science 230, 1165, 1985). How many properties or functions of red cells can one mimic synthetically? Can these synthetic cells serve as useful models? Here we report the first successful synthesis of an artificial model sickle cell. No reproducible, model cell system was previously available for research. A procedure identical to that used to prepare normal neohemocytes (NHC) was employed using sickle hemoglobin (HbS). The starting material was O2 or CO liganded HbS at a concentration of approximately 15g% in a 30 mOsm phosphate buffer; this solution was kept ultrahypotonic until the final stage of the process. The lipid bilayer membrane was formed during a prolonged adjustment of the osmolality to 300 mOsm. The final step was removal of unencapsulated HbS. Sickle NHC were examined in parallel with normal (HbA containing) NHC by scanning and thin section electron microscopy before and after deoxygenation. These synthetic cells do sickle! Some look remarkably like red blood cells, only much smaller. Our data suggests that polymerization of the HbS within sickle NHC may be initiated by a different mechanism than the polymerization of purified solutions of HbS. The typical lipid bilayer seen in HbA containing NBC was essentially absent in the sickle NHC: similar results have been reported for irreversibly sickled red cells. Sickle NHC thus have remarkable potential to function as model sickle cells.

About twenty years ago, The National Academy of Science-National Research Council's Committee on Plasma and Plasma Substitute, Division of Medical Sciences sponsored the Third Conference on Artificial Colloidal Agents. At that time there was some concern about the stability of dextran. Therefore there was the need for a more suitable artificial colloidal plasma volume expander. At this meeting the results were presented in order to evaluate the feasibility of using solutions of hydroxyethyl starch (HES). This polymer is a derivative of amylopectin, a “waxy” starch, containing only branched chains of glycoside molecules devoid of a linear component. It is prepared by first reacting the starch to acid hydrolysis and then with ethylene oxide in an alkaline solution. There are two distinct advantages to using HES as a plasma substitute. The first is the wide range of molecular weights that are available and the second is the degree of molar substitution, which is defined as the average nunber of hydroxyethyl groups reacted per anhydroglucose residue. It is this unique combination that is the major determinant for the rate of amylase hydrolysis and the time of retention. The half-life concentration parameter, that is, the time required for the concentration of HES in the bloodstream, following intravenous injection, to fall to 50 percent of its initial peak value, has a range of times from 24–36 hours for a weight average molecular weight, Mw, of 450,000 and degree of molar substitution, M.S., of 0.7 to 2–4 hours for a Mw of 40,000 and M.S. of 0.55, respectively [1].

PHP, mean molecular weight of approximately 90,000 daltons, is produced by chemical modifications of outdated human red cell hemoglobin (Hb). The P50 value of PHP is 22 ± 0.7mmHg. Eleven healthy mongrel dogs were exposed to severe anemia to a hematocrit of 5 ± 2% by normovolemic exchange (ET) with PHP (n=5) or a plasma expander, Hespan (Control, n=6). Hb concentration in the preparation is 6.0 g/dl. All the animals with PHP tolerated the procedure well and have survived for 1 year at sacrifice while five of the controls died within a week (p=0.004). Reductions in hematological and coagulation parameters occurred following ET and these parameters returned to the normal ranges by 4 weeks post ET. Renal function parameters (serum urea and creatinine) remained in the normal range throughout the experimental period. A transient slight increase in the hepatic enzyme SGOT was observed. Open biopsies of major organs at 2 weeks post ET showed vacuolated cells in the liver and kidneys. Normal histology was noted at and after 3 months. Oxygen transport properties as examined by arterial-venous oxygen differences, oxygen delivery and oxygen extraction showed that PHP continued to transport oxygen up to 48 hours studied post ET. PHP functioned as effectively as red blood cells at 6 hours in oxygen delivery to the tissues. PHP effectively supported life at lethal levels of anemia and is physiologically well accepted.

Stroma Free Hemoglobin (SFH) and Perfluorotributylamine (FC-43) have been extensively studied as oxygen carriers. Therefore, capitalizing on the role of oxygen reversing sickling, and thus decreasing resistance, this study was undertaken to investigate possible differences in the rheological effects of mixtures of FC-43 and SFH at different ratios. We used the positive pressure cell filtration system (PPCFS) of Lessin and Kurantsin- Mills. FC-43 was purchased from Alpha Therapeutics Corporation, and SFH was prepared by the method of Rabiner et al. The FC-43/SFH mixtures were made on a volume basis in ratios of 1:1 and 1:2. Fresh sickle cell anemia (HbSS) blood was used to prepare 0.2% cell suspensions which were deoxygenated, by bubbling with moist N2 for one hour. Studies of the rheological effects of FC-43/SFH mixtures on sickled erythrocytes were determined by obtaining filtration measurements on the following:- 1. sickle cells in ambient air; 2. deoxygenated sickle cells; 3. deoxygenated sickle cells + oxygenated FC-43/SFH (1:1 v/v); 4. deoxygenated sickle cells + oxygenated FC - 43/SFH (1:2 v/v). The results showed that treatment of deoxygenated sickle erythrocytes with oxygenated SFH/FC-43 mixtures decreased the flow resistance of the deoxygenated cells. Secondly, the SFH/FC-43 (1:2 v/v) mixture caused greater decrease in flow resistance than the 1:1 mixture.

Therefore it is reasonable to conclude that the mixtures of SFH/FC-43 (1:2 v/v) ratio appear to have a additive effect in reducing the flow resistance i.e. increasing the filterability of sickled erythrocytes.

Continuing refinement of fluorocarbon blood substitutes depends upon perfection and application of methods to monitor oxygen transport in living animals. Methods involving chronic catheterization, electrode implantation and tissue sampling yield valuable basic information in animal experiments. The importance of blood lactate in assessing the adequacy of fluorocarbon oxygen transport is reported here. The preliminary observation of microbubble blockade in the pulmonary capillaries in rabbits receiving F-decalin (PP5) emulsion, resulting in lungs which remain noncollapsible for at least 79 days is also reported.

The binding of propranolol by a perfluorochemical emulsion (PFCE), cXiacid glycoprotein (AGP), and their mixtures was examined by a dialysis exchange method at 37°C. A sedimentation method was used to verify propranolol binding by the PFCE. Binding of 100 and 500 ng/ml drug by 100, 75, 50 and 25% v/v concentrations of the PFCE or of 0.067% AGP was determined. Propranolol was highly bound by the PFCE with a percent free propranolol in undiluted PFCE of 13.1% at both 100 and 500 ng/ml propranolol. Propranolol was moderately bound by AGP. The percent free propranolol in 0.067% AGP was 31.4% and 32.4% for 100 and 500 ng/ml propranolol, respectively. The percent bound propranolol was not significantly different when the emulsion was diluted to 75 and 50% v/v with 0.067% AGP solution rather than buffer. When the PFCE was further diluted to 25% v/v with the 0.067% AGP solution the percent free propranolol was significantly lower than in the 25% v/v PFCE alone or 75% v/v protein solution alone. Thus, the PFCE has a relatively large affinity and capacity for propranolol. Propranolol did not partition from buffer to the PFC liquid. This indicates that propranolol bound by the PFCE is associated only with the emulsifiers.

The effect of a perfluorochemical blood substitute (Fluosol-DA 20%) on the disposition kinetics of digoxin in the dog was studied. An intravenous dose of digoxin was administered at either 0.25 or 24 hours following a 30% blood exchange with the dog's own blood (SHAM group) or Fluosol-DA (treatment group). A significant change in the initial volume of distribution and mean residence time in the serum was noted when digoxin was administered 0.25 hours after the Fluosol-DA blood exchange. This volume of distribution change was not observed when digoxin was administered 24 hours after the Fluosol-DA blood exchange. No change in the t1/2, total body clearance, or volume of distribution at steady-state was observed in any of the treatments. No change in the partitioning of digoxin in the blood was noted following Fluosol-DA administration.