Editor's Choice - The National Norwegian Carotid Study: Time from Symptom Onset to Surgery is too Long, Resulting in Additional Neurological Events.

OBJECTIVE/BACKGROUND: The objective was to observe for 1 year all patients in Norway operated on for symptomatic carotid stenosis with respect to (i) the time from the index event to surgery and neurological events during this time; (ii) the level in the healthcare system causing delay of surgical treatment; and (iii) the possible relationship between peri-operative use of platelet inhibitors and neurological events while awaiting surgery. METHODS: This was a prospective national multicentre study of a consecutive series of symptomatic patients. Patients were eligible for inclusion when referred for surgery. An index event was defined as the neurological event prompting contact with the healthcare system. All 15 departments in Norway performing carotid endarterectomy (CEA) participated. RESULTS: Three hundred and seventy one patients were eligible for inclusion between 1 April 2014 and 31 March 2015, and 368 patients (99.2%) were included. Fifty-four percent of the patients contacted their general practitioner on the day of the index event. Primary healthcare referred 84.2% of the patients to hospital on the same day as examined. In hospital median time from admission to referral for vascular surgery was 3 days. Median time between referral to the operating unit and actual CEA was 5 days. Overall, 61.7% of the patients were operated on within 2 weeks of the index event. Twelve patients (3.3%) suffered a new neurological event while awaiting surgery. The percentage of patients on dual antiplatelet therapy was lower (25.0%) in this group than among the other patients (62.6%) (p = .008). The combined 30 day mortality and stroke rate was 3.8%. CONCLUSION: This national study with almost complete inclusion and follow-up shows that the delays occur mainly at patient level and in hospital. The delay is associated with new neurological events. Dual antiplatelet therapy is associated with reduced risk of having a new neurological event before surgery.

HNA antibody-mediated neutrophil aggregation is dependent on serine protease activity.

BACKGROUND AND OBJECTIVES: Transfusion-related acute lung injury (TRALI) is often caused by antibodies against human neutrophil alloantigen-2 (HNA-2) and HNA-3a. Neutrophil aggregation is considered as a major cause of TRALI, but little is known about how HNA antibodies initiate this process. We explored mechanisms involved in neutrophil aggregation induced by HNA-2 and HNA-3a antibodies. MATERIALS AND METHODS: Isolated neutrophils were pretreated with broad-spectrum or specific inhibitors against different cell functions or proteases. Granulocyte agglutination test (GAT) was performed with serially diluted anti-HNA-2 and anti-HNA-3a plasmas or control plasma, and reactivity was evaluated microscopically. Reactive oxygen species (ROS) production in neutrophils was investigated using a lucigenin-based chemiluminescence assay. RESULTS: HNA-2 and HNA-3a antibody-mediated neutrophil aggregation was inhibited by pretreatment with formaldehyde, iodoacetamide and the serine protease inhibitors Pefabloc-SC, N-p-tosyl-L-phenylalanine chloromethyl ketone (TPCK) and Nα-tosyl-L-lysine chloromethyl ketone hydrochloride (TLCK). In contrast, inhibition of actin polymerization, respiratory burst, cysteine proteases, metalloproteases or aspartic proteases did not affect neutrophil aggregation. Furthermore, HNA-3a antibodies did not directly cause ROS production in neutrophils. CONCLUSION: Aggregation of neutrophils induced by HNA-2 and HNA-3a antibodies is an active process and depends on trypsin- or chymotrypsin-like serine proteases but is not dependent on the production of ROS. These findings may open new prospects for the pharmacologic prevention of neutrophil-associated acute lung injury.

Does the great saphenous vein stripping improve arterial leg blood flow during exercise?

OBJECTIVES: It has been shown that the leg muscle pump increases arterial leg blood flow during upright exercise in healthy subjects, and that this effect is reduced in patients with incompetence of the great saphenous vein (GSV). In this study, patients with GSV reflux causing varicose veins were investigated after GSV stripping, to see whether the muscle pump effect on arterial leg blood flow is improved. DESIGN: Prospective case study. METHODS: Nine patients with GSV incompetence resulting in symptomatic varicose veins, but without peripheral artery disease were included in this study. Patients exercised in the supine and 30° head up tilted positions by rhythmically pressing down a pedal with one foot. Blood flow was measured in the femoral artery using Doppler ultrasound. The Exercise-induced rise in femoral artery blood flow was compared in the supine and 30° head up tilted positions. Patients were investigated both before and after undergoing saphenofemoral ligation and GSV stripping as a treatment for their varicose veins. The arterial blood flow response to exercise was compared between the pre and postoperative observations. RESULTS: Prior to GSV stripping the immediate rise in femoral flow was 0.25 l min(-1) above rest in both supine and tilted positions. After GSV stripping however, the rise in flow was 30% larger in the tilted position than in the supine position (0.26 vs. 0.20 l min(-1), P < 0.05). CONCLUSIONS: GSV stripping modestly improves arterial leg blood flow at the onset of exercise in patients with GSV insufficiency, because of an improved effect of the leg muscle pump.

Platelets kill bacteria by bridging innate and adaptive immunity via platelet factor 4 and FcγRIIA.

Essentials Human platelets specifically interact with IgG opsonized bacteria through FcγRIIA. Platelet factor 4 (PF4) binds to polyanions (P) and undergoes a conformational change. Anti-PF4/P IgG opsonizes PF4-coated Gram-positive and Gram-negative bacteria. Platelets specifically kill E.coli opsonized with PF4 and human anti-PF4/P IgG. SUMMARY: Background Activated platelets release the chemokine platelet factor 4 (PF4) stored in their granules. PF4 binds to polyanions (P) on bacteria, undergoes a conformational change and exposes neoepitopes. These neoepitopes induce production of anti-PF4/P antibodies. As PF4 binds to a variety of bacteria, anti-PF4/P IgG can bind and opsonize several bacterial species. Objective Here we investigated whether platelets are able to kill bacteria directly after recognizing anti-PF4/P IgG opsonized bacteria in the presence of PF4 via their FcγRIIA. Methods Using platelet-bacteria suspension co-culture experiments and micropatterns with immobilized viable bacteria, in combination with pharmacological inhibitors and human anti- PF4/P IgG we analyzed the role of platelet-mediated killing of bacteria. Results In the presence of PF4, human anti-PF4/P IgG and platelets, E. coli killing (> 50%) with colony forming units (CFU mL-1 ) 0.71 × 104 ± 0.19 was observed compared with controls incubated only with anti-PF4/P IgG (CFU mL-1 3.4 × 104 ± 0.38). Blocking of platelet FcγRIIA using mAb IV.3 (CFU mL-1 2.5 × 104 ± 0.45), or integrin αIIbß3 (CFU mL-1 2.26 × 104 ± 0.31), or disruption of cytoskeletal functions (CFU mL-1 2.7 × 104 ± 0.4) markedly reduced E. coli killing by this mechanism. Our observation of E. coli killing by platelets on micropatterned arrays is compatible with the model that platelets kill bacteria by covering them, actively concentrating them into the area under their granulomere and then releasing antimicrobial substances of platelet α-granules site directed towards bacteria. Conclusion These findings collectively indicate that by bridging of innate and adaptive immune mechanisms, platelets and anti-PF4/polyanion antibodies cooperate in an antibacterial host response.

Uptake and intracellular transport of acidic fibroblast growth factor: evidence for free and cytoskeleton-anchored fibroblast growth factor receptors.

Endocytic uptake and intracellular transport of acidic FGF was studied in cells transfected with FGF receptor 4 (FGFR4). Acidification of the cytosol to block endocytic uptake from coated pits did not inhibit endocytosis of the growth factor in COS cells transfected with FGFR4, indicating that it is to a large extent taken up by an alternative endocytic pathway. Fractionation of the cells demonstrated that part of the growth factor receptor was present in a low-density, caveolin-containing fraction, but we were unable to demonstrate binding to caveolin in immunoprecipitation studies. Upon treatment of the cells with acidic FGF, the activated receptor, together with the growth factor, moved to a juxtanuclear compartment, which was identified as the recycling endosome compartment. When the cells were lysed with Triton X-100, 3-([3-chloramidopropyl]dimethylammonio)-2-hydroxy-1-propanesulfona te, or 2-octyl glucoside, almost all surface-exposed and endocytosed FGFR4 was solubilized, but only a minor fraction of the total FGFR4 in the cells was found in the soluble fraction. The data indicate that the major part of FGFR4 is anchored to detergent-insoluble structures, presumably cytoskeletal elements associated with the recycling endosome compartment.

Platelet activation in the presence of neutral protamine Hagedorn insulin: a new feature of antibodies against protamine/heparin complexes.

Essentials Protamine (PRT) is used to stabilize insulin in neutral protamine Hagedorn (NPH) insulin. The interaction between NPH-insulin, anti-PRT/heparin antibodies and platelets was investigated. Anti-PRT/heparin antibodies activate platelets in presence of NPH-insulin dependent on heparin. Cross-reactivity seems to have no major effect on the clinical outcome of medical patients. SUMMARY: Background Protamine (PRT) is used to stabilize insulin in neutral protamine Hagedorn (NPH) insulin, a commonly used therapeutic agent for diabetes mellitus. Immunization against PRT/heparin complexes is common in diabetic patients. Objectives To investigate the impact of NPH-insulin on the interaction between anti-PRT/heparin antibodies and platelets. Methods The interaction between NPH-insulin and anti-PRT/heparin antibodies was tested using in-house enzyme immunoassays. The ability of anti-PRT/heparin antibodies to activate platelets in the presence of NPH-insulin (and heparin) was investigated using flow cytometry. Results Twenty-one out of 80 sera containing anti-PRT/heparin IgG showed binding to NPH-insulin. Anti-PRT/heparin IgG from immunized patients bound to platelets in the presence of NPH-insulin, but not in the presence of native insulin. Anti-PRT/heparin antibodies induced P-selectin expression in the presence of NPH-insulin in a heparin-dependent way (median mean fluorescence intensity in the presence of NPH-insulin: 55, 95% confidence interval [CI] 18.7-100.5 vs. NPH-insulin and heparin: 204, 95% CI 106.5-372.8). The clinical relevance of platelet-activating anti-PRT/heparin antibodies was assessed by investigating a multicenter study cohort of 332 acutely ill medical patients who received heparin. None of the 21 patients with anti-PRT/heparin IgG developed thrombocytopenia or thromboembolic complications. Conclusions Anti-PRT/heparin antibodies activate platelets in the presence of NPH-insulin in a heparin-dependent way. However, results from our preliminary study indicate no major impact of these antibodies on the clinical outcome in medical patients receiving heparin, particularly on thromboembolic complications.

Encapsulation of islets in rough surface, hydroxymethylated polysulfone capillaries stimulates VEGF release and promotes vascularization after transplantation.

The transplantation of encapsulated islets of Langerhans is one approach to treat type 1 diabetes without the need of lifelong immunosuppression. Capillaries have been used for macroencapsulation because they have a favorable surface-to-volume ratio and because they can be refilled. It is unclear at present whether the outer surface of such capillaries should be smooth to prevent, or rough to promote, cell adhesions. In this study we tested a new capillary made of modified polysulfone (MWCO: 50 kDa) with a rough, open-porous outer surface for islet transplantation. Compared with free-floating islets, encapsulation of freshly isolated rat islets affected neither the kinetics nor the efficiency of glucose-induced insulin release in perifusion experiments. Free-floating islets maintained insulin secretion during cell culture but encapsulated islets gradually lost their glucose responsiveness and released VEGF. This indicated hypoxia in the capillary lumen. Transplantation of encapsulated rat islets into diabetic rats significantly reduced blood glucose concentrations from the first week of implantation. This hypoglycaemic effect persisted until explantation 4 weeks later. Transplantation of encapsulated porcine islets into diabetic rats reduced blood glucose concentrations depending on the islet purity. With semipurified islets a transient reduction of blood glucose concentrations was observed (2, 8, 18, 18 days) whereas with highly purified islets a sustained normoglycaemia was achieved (more than 28 days). Explanted capillaries containing rat islets were covered with blood vessels. Vascularization was also observed on capillaries containing porcine islets that were explanted from normoglycaemic rats. In contrast, on capillaries containing porcine islets that were explanted from hyperglycemic rats a fibrous capsule and lymphocyte accumulations were observed. No vascularization on the surface of transplanted capillaries was observed in the absence of islets. In conclusion, encapsulated islets can release VEGF, which appears to be an important signal for the vascularization of the capillary material. The rough, open-porous outer surface of the polysulfone capillary provides a site well suited for vascular tissue formation and may allow a prolonged islet function after transplantation.

In vitro test of new biomaterials for the development of a bioartificial pancreas.

The implantation of macroencapsulated islets has the potential to restore endogenous insulin secretion in type 1 diabetics, with no need for lifetime immunosuppression. To match the physiological fluctuations of blood glucose concentrations with appropriate insulin release, the macroencapsulation material must combine immunoprotection with optimal diffusion properties for glucose and insulin. The impact of chemical modifications of polysulphone (PSU) capillary polymers with a cutoff of 50 kD on glucose-induced insulin secretion of macroencapsulated rat islets was studied in perifusion experiments. The insulin release of free-floating islets showed the typical rapid response to glucose stimulation. Total insulin release (AUC between minute 30 and 120 of perifusion) reached 117+/-22 ng/ml. Blending PSU with polyvinylpyrrolidone or sodium-dodecyl-sulfate was not suitable for islet macroencapsulation, since glucose-induced insulin release was absent or disturbed. Hydroxy-methylation (CH2OH) of PSU improved the secretory behavior of macroencapsulated islets depending on the degree of PSU substitution (DS 0.8, AUC 62+/-15 ng/ml; DS 1.8, 111+/-24 ng/ml). In highly substituted PSU-capillaries the kinetics of glucose-induced insulin release was very similar to that observed in free-floating islets. Two consecutive glucose stimulations potentiated insulin release of free-floating islets during the second period of stimulation. Furthermore, freshly isolated macroencapsulated islets responded with more efficient insulin secretion after the initial priming. In conclusion, in vitro membrane screening identified highly substituted hydroxy-methylated PSU as the material of choice for islet encapsulation in a bioartificial pancreas.

Gradual increase in leg oxygen uptake during repeated submaximal contractions in humans.

We examine whether muscle oxygen consumption (VO2) increases gradually during repeated submaximal isometric contractions. Six subjects made two-legged isometric quadriceps contractions at 30% maximal voluntary contraction for 6 s with 4 s of rest between until exhaustion (58 +/- 8 min). Blood samples were taken from the femoral vein and artery, and blood velocity was recorded by ultrasound-Doppler technique in the femoral artery. Blood flow was calculated from velocity and artery diameter values. Leg VO2 increased sixfold within the 1st min of exercise. A further doubling of the VO2 was seen during the remainder of the exercise, reaching 307 +/- 22 ml/min at exhaustion. This latter increase was due to a 54% increase in blood flow and a 34% increase in oxygen extraction. After 20 min of recovery VO2 was still 75% higher than preexercise values. The results show a twofold increase in energy demand of the working muscle during repeated constant-force isometric contractions. The increased energy cost of contraction is probably localized at the cellular level, and it parallels fatigue determined as decreased force-generating capacity.

Areal density measurement is a convenient method for the determination of porcine islet equivalents without counting and sizing individual islets.

The determination of islet mass is important for the normalization of islet experiments in the laboratory and for the precise dosing of islets for transplantation. The common microscopical analysis is based on individual islet sizing, calculation of the frequency distribution, and conversion into islet equivalents (IEQ), which is the volume of a spherical islet with a diameter of 150 microm. However, islets are of irregular form, which makes this determination user dependent, and the analysis is irreproducible once the original sample is discarded. This routine technique of islet quantification was compared with the analysis of areal density measurements. It was assumed that the entire area occupied by islets can be expressed in IEQ without sizing and counting individual islets. Porcine islets were isolated by continuous digestion/filtration and purified by gradient centrifugation. Purified islets were stained with dithizone and were repeatedly pictured under the microscope with random area selection. A total of 51 pictures was taken from 11 different purifications and stained islets were detected by digital image analysis. The correlation coefficient (r) between bothanalyses was 0.977 with an underestimation of islet yield by areal density detection (slope: 0.75 +/- 0.03). Areal density analysis per picture took about 1 min, which is about 10 times faster than the traditional method without increasing the method error (CV 2.1% vs. 2.7%). In summary, areal density measurements allow a rapid and reproducible estimation of IEQ without counting individual islets. It can be performed in a single step analysis without computer programming and is valuable for online determinations of islet yield preceding transplantation.

Haemodynamic contributions to post-exercise hypotension in young adults with hypertension and rapid resting heart rates.

The haemodynamic effects of 45 min of treadmill exercise (at 70% of resting heart rate reserve) were determined in 5 young adults with hypertension and rapid resting heart rates (greater than 90 beats/min in clinic) and were compared with those of 5 age-matched normotensive subjects. Blood pressure was lower after exercise in the hypertensive, but not the normotensive subjects. Mean cardiac output before exercise was similar in the two groups, and fell from 6.8 +/- 0.6 before to 5.4 +/- 0.6 l/min 60 min after exercise in the hypertensive group (P less than 0.01). Total peripheral resistance tended to be higher at this time. Neither variable was affected by prior exercise in the normotensive group. The depressor effects of prior exercise on mean arterial pressure (-8.6 +/- 1.0 vs. -1.4 +/- 2.5 mmHg; P less than 0.04) and cardiac output (-1.4 +/- 0.3 vs. -0.1 +/- 0.1 l/min; P less than 0.005) and the increase in total peripheral resistance (+3.0 +/- 1.2 vs. 0.0 +/- 1.0 Units; P less than 0.05) were greater in the hypertensive group. Thus, the post-exercise hypotension in this selected group of young hypertensive subjects with rapid resting heart rates was mediated by a decrease in cardiac output and stroke volume disproportionate to the fall in blood pressure, suggesting sustained compromise of their cardiac performance after acute exercise.

Macroencapsulation of rat islets without alteration of insulin secretion kinetics.

Transplantation of encapsulated islets may restore endogenous insulin secretion in type 1 diabetics with no need of lifetime immunosuppression of the recipient. A biomaterial should be developed which combined immunoisolation with rapid and efficient diffusion of glucose and insulin. Rat islets were macroencapsulated in capillaries (molecular cut off 50 kD) of differently modified polysulphone. Macroencapsulated islets were perifused to study the kinetics of glucose induced insulin secretion into the perifusion medium. Blending polysulphone (PSU) with poly vinyl pyrrolidone or sodium dodecyl sulphate was not suited for islet macroencapsulation since glucose induced insulin release was absent after encapsulation. Hydroxy methylation (CH2OH) of PSU improved the secretory behaviour of macroencapsulated islets depending on the degree of substitution (DS). At 0.8 DS glucose induced insulin secretion was delayed and inefficient. At maximal degrees of PSU-substitution (1.8) the kinetics of insulin release and the efficiency of insulin release were very similar to that observed of free floating islets. In conclusion, highly substituted hydroxy methylated polysulphone allows a rapid and efficient insulin release after macroencapsulation and is suited for the further development of a bioartificial pancreas.

The impact of the Trial Coordinator in the Diabetes Control and Complications Trial (DCCT). The DCCT Research Group.

The Diabetes Control and Complications Trial (DCCT) is a multicenter, randomized clinical trial studying the effects of two different diabetes regimens on the development and progression of early vascular complications in persons with insulin-dependent diabetes mellitus (IDDM). All of the centers have a Trial Coordinator. We administered a self-report questionnaire to each center to document the different activities for which the Trial Coordinator assumed responsibility in successfully orchestrating the trial. All Trial Coordinators were responsible primarily for recruitment, screening, medical management, education and training, and adherence and administration. Although documentation indicated that the Trial Coordinator was responsible for all of the above activities, the original applications reflected that very few of the Principal Investigators anticipated such a wide variety of duties. A Trial Coordinator was named in only 13 of the 21 applications and of these, only 6 actually assumed the position. This study points out the need to develop a means to define characteristics, background, and training appropriate for candidates for a Trial Coordinator position in future studies.

Dentists and mandated reporting of child abuse and neglect.

Dentists rarely report child abuse and neglect. The author discusses dentists' failure to report and puts it in perspective by surveying the literature and conducting original research. He also offers solutions to assist dentists in fulfilling their responsibilities to recognize and report maltreatment. Dentists can become part of the child abuse and neglect service system, which will help abused and neglected children and keep dentists from legal and financial liability.

RSK2 regulates endocytosis of FGF receptor 1 by phosphorylation on serine 789.

FGFR1 (fibroblast growth factor receptor 1) regulates many key cellular responses including proliferation, migration and differentiation through activation of signaling pathways. Irregularities in FGFR1 signaling have been implicated in several pathological conditions, including human cancer. In order to discover novel regulators of FGFR1 signaling, we performed yeast two-hybrid screens and identified RSK2 (p90 ribosomal S6 kinase 2) as a potential FGFR1 interaction partner. RSK2 belongs to the family of serine/threonine kinases that are activated through the Ras-MAPK signal transduction pathway. Both in vitro and in vivo experiments confirmed the interaction and we show that phosphorylated RSK2 binds to and phosphorylates serine 789 in the C-terminal tail of FGFR1. Inhibition of RSK2 activity led to prolonged tyrosine transphosphorylation of FGFR1. Furthermore, prevention of FGFR1 phosphorylation by inhibition of RSK2 activity or mutation of serine 789 to alanine reduced FGFR1 endocytosis and ubiquitination explaining mechanistically the prolonged signaling activity. We propose a novel regulatory mechanism whereby activated RSK2 directly interacts with and phosphorylates FGFR1, thereby modulating receptor signaling through regulation of endocytosis.

Does venous insufficiency impair the exercise-induced rise in arterial leg blood flow?

OBJECTIVES: It has been shown that the leg muscle pump increases the immediate rise in arterial leg blood flow during upright exercise in healthy subjects. The present study is the first to investigate the muscle pump effect in exercise hyperaemia in patients with venous insufficiency, who should be lacking an optimally functioning muscle pump. METHODS: Any muscle pump effect is more pronounced in an upright position because of gravitation. The exercise-induced rise in femoral artery flow (FF) (ultrasound Doppler) was thus compared in the supine and 30° head-up tilted position in 10 patients. RESULTS: Neither the transient nor the steady-state rise in FF showed any difference between positions. This is in contrast to the previous findings in healthy subjects, where the transient rise in FF was larger in the tilted position. CONCLUSION: The muscle pump effect in exercise hyperaemia seems to be reduced or lacking in these patients.

The bphC gene-encoded 2,3-dihydroxybiphenyl-1,2-dioxygenase is involved in complete degradation of dibenzofuran by the biphenyl-degrading bacterium Ralstonia sp. SBUG 290.

AIMS: Biphenyl-degrading bacteria are able to metabolize dibenzofuran via lateral dioxygenation and meta-cleavage of the dihydroxylated dibenzofuran produced. This degradation was considered to be incomplete because accumulation of a yellow-orange ring-cleavage product was observed. In this study, we want to characterize the 1,2-dihydroxydibenzofuran cleaving enzyme which is involved in dibenzofuran degradation in the bacterium Ralstonia sp. SBUG 290. METHODS AND RESULTS: In this strain, complete degradation of dibenzofuran was observed after cultivation on biphenyl. The enzyme shows a wide substrate utilization spectrum, including 1,2-dihydroxydibenzofuran, 2,3-dihydroxybiphenyl, 1,2-dihydroxynaphthalene, 3- and 4-methylcatechol and catechol. MALDI-TOF analysis of the protein revealed a strong homology to the bphC gene products. We therefore cloned a 3.2 kb DNA fragment containing the bphC gene of Ralstonia sp. SBUG 290. The deduced amino acid sequence of bphC is identical to that of the corresponding gene in Pseudomonas sp. KKS102. The bphC gene was expressed in Escherichia coli and the meta-fission activity was detected using either 2,3-dihydroxybiphenyl or 1,2-dihydroxydibenzofuran as substrate. CONCLUSIONS: These results demonstrate that complete degradation of dibenzofuran by biphenyl degraders can occur after initial oxidation steps catalysed by gene products encoded by the bph-operon. The ring fission of 1,2-dihydroxydibenzofuran is catalysed by BphC. Differences found in the metabolism of the ring fission product of dibenzofuran among biphenyl degrading bacteria are assumed to be caused by different substrate specificities of BphD. SIGNIFICANCE AND IMPACT OF THE STUDY: This study shows for the first time that the gene products of the bph-operon are involved in the mineralization of dibenzofuran in biphenyl degrading bacteria.

Encapsulation of Islets in Rough Surface, Hydroxymethylated Polysulfone Capillaries Stimulates VEGF Release and Promotes Vascularization after Transplantation.

The transplantation of encapsulated islets of Langerhans is one approach to treat type 1 diabetes without the need of lifelong immunosuppression. Capillaries have been used for macroencapsulation because they have a favorable surface-to-volume ratio and because they can be refilled. It is unclear at present whether the outer surface of such capillaries should be smooth to prevent, or rough to promote, cell adhesions. In this study we tested a new capillary made of modified polysulfone (MWCO: 50 kDa) with a rough, open-porous outer surface for islet transplantation. Compared with free-floating islets, encapsulation of freshly isolated rat islets affected neither the kinetics nor the efficiency of glucose-induced insulin release in perifusion experiments. Free-floating islets maintained insulin secretion during cell culture but encapsulated islets gradually lost their glucose responsiveness and released VEGF. This indicated hypoxia in the capillary lumen. Transplantation of encapsulated rat islets into diabetic rats significantly reduced blood glucose concentrations from the first week of implantation. This hypoglycaemic effect persisted until explantation 4 weeks later. Transplantation of encapsulated porcine islets into diabetic rats reduced blood glucose concentrations depending on the islet purity. With semipurified islets a transient reduction of blood glucose concentrations was observed (2, 8, 18, 18 days) whereas with highly purified islets a sustained normoglycaemia was achieved (more than 28 days). Explanted capillaries containing rat islets were covered with blood vessels. Vascularization was also observed on capillaries containing porcine islets that were explanted from normoglycaemic rats. In contrast, on capillaries containing porcine islets that were explanted from hyperglycemic rats a fibrous capsule and lymphocyte accumulations were observed. No vascularization on the surface of transplanted capillaries was observed in the absence of islets. In conclusion, encapsulated islets can release VEGF, which appears to be an important signal for the vascularization of the capillary material. The rough, open-porous outer surface of the polysulfone capillary provides a site well suited for vascular tissue formation and may allow a prolonged islet function after transplantation.