Tumor necrosis factorr(-T-NEra)._and interleukin-1 (IL-lp) are cytokines primarily produced by monocytes/macrophages when stimulated by endotoxin, complementderived anaphylatoxins and the specific antigen. In the present study, the plasma levels of TNF-a and IL-lp were evaluated before and after hemodialysis with cuprophane membrane (in 9 patients) and hemodiafiltration (in 9 patients) using three high-permeability membranes such as polymethylmethacrylate, polyacrylonitrile (AN-69) and polysulfone In vitro spontaneous production ofTNF-a and IL-lp was evaluated in the supernatants from short-term cultured monocytes obtained before and after treatment. The predialytic levels ofTNF-a and IL-lp were significantly higher (p < 0.05) in the uremic population than in 1 healthy subjects taken as controls. The analysis**the uremic population regarding the mode of therapy indicated that in hemodialysis thepredialytic plasma levels ofTNF-a and IL-ip did not significantly differ from those of healthy subjects. In contrast, in hemodiafiltration with polymethylmethacrylate and AN-69, but not with polysulfone, the predialytic plasma levels of both cytokines were significantly (p < 0.05) increased. No significant variation in plasma levels of both cytokines was observed after hemodialysis with cuprophane membranes. Hemodiafiltration with polymethylmethacrylate and AN-69 but not with polysulfone, brought about a consistent reduction in plasma levels of both cytokines Detectable amounts ofTNF-a and IL-ip were spontaneously produced by peripheral-blood monocytes 6 h after the end of hemodialysis but not of hemodiafiltration. These studies suggest a possible role ofTNF-a and IL-ip in the biocompatibility of different extracorporeal treatments. 338 Tetta/Camussi/Turello/Salomone/Aimo/Priolo/Scgoloni/Vercellonc Introduction relevance in at least some of the acute and chronic clinical symptoms and biohumoral Intcrlcukin-ip(IL-ip)andtumornecrosis alterations observed in hemodialysis pafaclor-a (TNF-a) participate in the so-called tientssuch as fever, hypotension, low plasma polypeptide mediator network, which dezinc levels, increased plasma levels ofacutefincs a complex array of interacting signals pliase response proteins and late occurrence that regulates growth, differentiation and of amyloidosis. However, only few studies function of the cells involved in immunity, have so farexamined the in vivo production hematopoiesis and inflammation [1]. IL-ip ofTNF-a in hemodialysis patients, and TNF-a are produced primarily by The present study was undertaken to mononuclearphagocyteswhenthesecellsare evaluate in patients on hemodialysis with stimulated by a variety of agents including complement-bioincompatible cuprophane microbes, microbial products, inflammatory and in patients on hemodiafiltration with agents such as C5a, C3a and their carboxyhigh-permeability membranes [polymethylpeptidase-B-derived des-arginated products, methacrylate, polyacrylonitrile (AN-69), plant lectins and antigens [2, 3]. Furtherpolysulfone] their preand postdialytic more, TNF-a is a stimulus forIL-ip producplasma levels and in vitro spontaneous protion from blood monocytesand vascularen: duction of TNF-a and IL-ip from monodothelial cells [4, 5]. Endotoxin is the most cytes in short-term culture, potent soluble inducer of IL-ip and TNF-a The results ofthe present study suggest an production in vitro requiringconcentrations in vivo production of IL-ip and TNF-a in of picograms to nanograms per milliliter. the interdialytic period in patients on heTNF-aplaysaprimaryroleinthealterations modiafiltration with high-permeability associated with endotoxemia and septic membranes as inferred from the increased shock. IL-ip and TNF-a are recognized as predialytic plasma levels of both cytokines. potent mediators of inflammation [6-9]. However, hemodiafiltration with polymethConsiderable growth of organisms in the ylmethacrylate and AN-69, but not polysuldialysis fluid and high levels of endotoxin fone, drastically reduces the plasma levels of especially when bicarbonate concentrate is both cytokines. Interdialytic production of used as dialysate buffer are known to occur IL-ip and TNF-a is markedly reduced in and have been implicated in the induction of '* patients on hemodialysis with cuprophane. synthesis and release ofIL-1p in in vitro and clinical hemodialysis [reviewed in ref. 2, 10]. Microbial contamination ofthe dialysis fluid Materials and Methods has drawn increasing interest since backfilp. tration ofdialysate into the blood compart^ZL chronically uremic patients entered this ment indeed occurs with today's high-perstudy.All patientsreported no febrilereaction during meability membranes [reviewed in ref.. 10]. ora^tcr hemodialysis or at home and they had been As first outlined by Henderson et al. [1 1] and 'freeofintercurrentd«sease forthe last 4 weeks. ExcluShaldon et al. [12], several of the overlapsloncriteria were: PreseiU orPast historv of diabetes; ,. , . , A. . ' .-_ tn , ,,.. autoimmunediseases;Australiaantigenserumposipmgbiological activities ofIL-ip andTNF-a tivity. intolerance t0 lhe hemodialysis membranes in [reviewed in ref. 1 3] may.be of pathogenetic study; therapy with corticosteroids; fi-receptor-block3 39 !' Cytokines in Hemodialysis __ ing agents, and anti-inflammatory, immunosupprcssis with cuprophane membranes and of Patents on sive or fat-lowering drugs. The primary renal disease hemodiafiltration with polymethylmethacrylate (9), waspolycystickidneydiseasein 2 patients, multicystic AN-69 (9) and polysulfonc (9). Separation of monokidney in I,congenital malformation in I,chronicglonuclearcells was achieved afterccnlnfugalion (700^ merulonephritis in 2, ncphroangiosclcrosis in 8, 20 min) using sterile lymphocyte separator tubes chronicpyelonephritisin 3 and Hcnoch-Schonleindis(FAR Italia, Verona, Italy) and undiluted Lymphocasc in 1. Patients werecategorized in twogroupson prep (Nycomcd. Oslo, Norway). The light density the basis of the hemodialysis technique used: 9 pafraction contained mainly lymphocytes and 10/o tients(meanage 59.3 ± 13.7years, meandialyticage monocytes. After 3 washes in RPMI 1640 (Gruppo 9 5 ± 6 5 years) were on hemodialysis with cuproFlow, Opera, Italy), the cells were allowed to adhere phane (Bravo 501), and 9 patients (mean age 53.8 ± (I h,37'C, 5%CO2),in6-wellplatesattheconcentra19.1 years, mean dialyticage 6.1 ± 4.3 years)undertionof2 X 10'cells/mlofRPMI 1640supp emen ed went hemodiafiltration, a simultaneous application of with 10% fetal calfscrum (G.bco, Paisley, UK), 100 hcmofiltration to hemodialysis [14, 15], using polyIU/mlpenicillinand 100ug/mlstreptomycin.Nonadmethylmethacrylate(Bl-1.6U, Hoechst-Toray, Kyoto, herent cells were eliminated by 3 washes and the . Japan) AN-69 (Filtral 16, Hospal) and polysulfone adherent cells were identified as monocytes (90 /o . (BL632 Bellco) The order of high-flux membranes pure) on the basis oftheir staining with nonspecific was randomized. The apparatus used was a Hospal cstcrase, usinga-naphthol-AS-D-chloroacetate as subhemodiafiltration system, with an ultrafiltration rate strate, combined with staining abrogation with socontroller and Equalizer. The substitution fluid was dium fluoride. RPMI contained no detectable endoinfused into avenouslinetocompensateforwaterand toxin as assessed by the Limulus assay Spontaneous solute removal (after dilution) and had the following releaseofTNF-oand IL-lp was assessed in the supercomposition (in mEq/1): Na* 140; K+ 1.5; Ca** 4.5; natants of adherent monocytes after 6 h culture. Mg2+ I; Cl105; lactate 42; glucose (5.5 mg%); osmolarity (297 mosm/1), pH 5-6. Dialysate was counterTNF-a Assay currentlydeliveredsimultaneouslyataflowrateof500 Asensitivebiological assay was used to quantitate ml/min Ultrafiltrate removed during each treatment TNF-a in plasma and in the supcrnatants of cultured ranged from 9-10 liters,which wasapproximately 2-3. monocytes that was based on its cytotox.c activity in liters morethanthe infused substitution fluid. Pre-and'. the presence ofan inhibitor of protein synthesis [16]. postdialytic hematocrit was evaluated. All patients Two-fold serial dilutions of samples were added towere dialyzed with the same membrane for 2 running getherwith 0.1 mg/ml ofcycloheximideto cultures of sessions before the study. Blood (5 ml) was drawn in human (SK-MEL-109) melanoma cells that are sensisterile heparinized glass tubes, centrifuged and the tive to the cytotoxic activity ofTNF-a in concentraplasma was aspirated, filtered sterile (Acrodisc, 0.2 tions as low as 20 pg/ml. These cells were grown as um)andstoredat4-CbeforeTNF-aand IL-lpassays. monolayers in 24-well cluster plates, incubated with SamplesfordeterminationofplasmaC3adesArgwere appropriate dilutions of samples and after 20h drawn in Na-EDTA glass tubes and immmediately washed with phosphate-buffered saline before staincentrifuged(700g, 15 min,4à"C)attime0, 15 minand ingwith crystalvioletwhichwaseluted and measured at the end of treatment. The plasma was recentrifuged as described earlier [16]. A calibration curve was confor 5 min at 4°C in a Beckman Microcentrifuge B structedwithhumanrecombinantTNF-a(0.01,0.1, I (Beckman Analytical, Milano, Italy) and stored at ng/ml) to convert the cytotoxic activity ^biological -80°C until assayed for C3a des Argby radioimmusamples into nanograms per m.ll.liter ofTNF-a. In noassay (Amity, Milano, Italy). order to assess specificity forTNF-a-induced cytotoxControl subjects were 21 healthy members of our icity, samples which exerted > 30% cytotox.c.ty on institution (age range 32-42 years). the TNF-a-scnsitive SK-MEL-109 cell line were assayed on TNF-a-resistant cells (designated R4) sePrcparation of Peripheral-Blood Monocytes lected from SK-MEL-1 09 cells. The lack of cytotoxicPeripheral-blood monocytes were purified from ity ofTNF-a-resistant R4 cells in such control experpreand postdialysis heparinized blood drawn from iments s

In summary, Canada enjoys a publicly funded healthcare system which is the second most costly in the world. The treatment rate for renal replacement therapy is comparable with countries in Europe but falls short of that in the United States and Japan. With respect to costs, peritoneal dialysis is a cheaper option than haemodialysis. The major costs in peritoneal dialysis are supply related and the major costs in haemodialysis are labour and supplies. A treatment modality change includes a set of costs which makes the most cost-effective therapy in almost cases a therapy that can be successfully maintained.

Between January 1, 1970, and December 31, 1994, 1,926 cases of biopsy-proven primary glomerulonephritis (PGN) were diagnosed in an adult population (>15 years of age) in a northwestern region of Italy with approximately 3.5 million inhabitants. The principal long-term changes were an increase in the absolute number of biopsies per year, an increase in the mean age of patients undergoing biopsy (from 29.3 ± 12.2 years to 47.0 ± 17.8 years), an increase in the percentage of patients older than 65 years (from 1.7% to 20.4%), and an increase in the percentage of isolated urinary abnormalities as an indication for biopsy (from 3.5% to 29.6%). In the total biopsy material, immunoglobulin A glomerulonephritis (IgA-GN) is the most frequent type (26%), followed by membranous glomerulonephritis (MGN; 20%). An incidence study was begun in 1990; this survey was restricted to the population of the province of Torino (approximately 2 million inhabitants) as only this area completely refers to the nephrologic centers that entered patients into this study. The overall incidence of PGN is 4.68 new cases/yr/105 population with a predominance of males (>2:1); IgA-GN is the most common type (1.47/yr/105 population [34.5%]) in the overall population. In the elderly, cases of PGN are twice as high as in adults (8.19/yr/105 population v 4.02/yr/105 population in the 65 to 74 year and 45 to 54 year age groups, respectively); MGN mainly accounts for this high incidence (3.4/yr/105population), while the nephrotic syndrome is the most common indication for biopsy (53.8%). A comparison with the incidence in the same area in the early 1970s is evaluable only for PGN, which was mainly registered in the age groups for which an unrestricted biopsy policy was already in place (15 to 35 years). In contrast with a misleading increase of all types of PGN, which is in reality due to the extension of the biopsy policy to older and asymptomatic patients, membranoproliferative glomerulonephritis type I shows a countercurrent decrease from 0.43 to 0.13/yr/105 population. Evidence of a simultaneous decrease in severe cardiac valvulopathy, due to rheumatic fever, is also provided. We feel that before epidemiologic conclusions can be reached, a clear understanding of one's own biopsy policy is essential. An apparent change in the PGN rate in our region over the last 25 years mainly depends on modifications in our biopsy policy, most probably coupled with a change in the threshold of detection of symptoms in the general population. At present, according to our experience, IgA-GN is the most common type of PGN in the total bioptic material, as demonstrated in other European countries, while the elderly show a peculiar pattern with a higher PGN incidence, mainly represented by MGN and heralded by the nephrotic syndrome. We also confirm that membranoproliferative glomerulonephritis type I is indeed decreasing in parallel with changes in the microbiologic environment.

A regular dialytic treatment of diabetic patients is until accepted from about twenty years in many areas. Aim of this work was a retrospective analysis of main clinical and survival data of diabetic patients (diabetic nephropathy or diabetes as comorbidous factor = 659 cases) admitted for dialysis in Piedmont (Northern Italy Region about 4,400,000 inhabitants) in the period 1981-1993 (functional recovery and follow-up < 1 month excluded). A progressive increment in incidence of diabetic patients was seen mostly in the aged. At 12/31/1993, 263 of 2404 patients admitted for dialysis were diabetics (10.9%); the majority of them was treated in Hospital Centers with bicarbonate haemodialysis (54.4%), while a small group was treated with CAPD (12.9%). During the years ¿80 was seen a progressive leaving of CAPD as first choice method in this population and in the last period the orienteering is the utilization of mixed methods (diffusive-convective as first choice). As regards the survival are not prominent significant differences between this cohort and the cohort affected by vasculopathy as comorbidous factor (86.2 and 54.2% in diabetics vs 78.6 and 55.2% in patients affected by vasculopathy at 1 and 3 years--p = 0.3481; patients aged 45-64 years). In conclusion the cohort of diabetic patients represent a good marker of the clinical problems of the elder population with high clinic risk, in progressive increasing in our Region.

It is well recognized that organ transplantation is associated with an increased number of tumors. Cutaneous malignancies are reported to be the most frequent in renal allograft recipients , in papers from Europe, USA and Australia. Their rate increases with duration of the graft, so that the cumulative risk rises to 40

There is increasing evidence that the biochemical and cellular phenomena induced by blood/ membrane/dialysate interactions contribute to dialysis-related intradialytic and long-term complications. However, there is a lack of large, prospective, randomized trials comparing biocompatible and bioincompatible membranes, and convective and diffusive treatment modalities. The primary aim of this prospective, randomized trial was to evaluate whether the use of polysulfone membrane with bicarbonate dialysate offers any advantage (in terms of treatment tolerance, nutritional parameters and pre-treatment beta-microglobulin levels) over a traditional membrane (Cuprophan). A secondary aim was to assess whether the use of more sophisticated methods consisting of a biocompatible synthetic membrane with different hydraulic permeability at different ultrafiltration rate (high-flux hemodialysis and hemodiafiltration) offers any further advantages. Seventy-one Centers were involved and stratified according to the availability of only the first two or all four of the following techniques: Cuprophan hemodialysis (Cu-HD), low-flux polysulfone hemodialysis (LfPS-HD), high-flux polysulfone high-flux hemodialysis (HfPS-HD), and high-flux polysulfone hemodiafiltration (HfPS-HDF). The 380 eligible patients were randomized to one of the two or four treatments (132 to Cu-HD, 147 to LfPS-HD, 51 to HfPS-HD and 50 to HfPS-HDF). The follow-up was 24 months. No statistical difference was observed in the algebraic sum of the end points between bicarbonate dialysis with Cuprophan or with low-flux polysulfone, or among the four dialysis methods under evaluation. There was a significant decrease in pre-dialysis plasma beta 2-microglobulin levels in high-flux dialysis of 9.04 +/- 10.46 mg/liter (23%) and in hemodiafiltration of 6.35 +/- 12.28 mg/liter (16%), both using high-flux polysulfone membrane in comparison with Cuprophan and low-flux polysulfone membranes (P = 0.032). The significant decrease in pre-dialysis plasma beta 2-microglobulin levels could have a clinical impact when one considers that beta 2-microglobulin accumulation and amyloidosis are important long-term dialysis-related complications.

The aim of the present study was to investigate the angiogenic properties of platelet-activating factor (PAF). In vitro PAF was shown to induce a dose-dependent migration of human endothelial cells (EC) across the polycarbonate filters in Boyden's chambers. In contrast, D-PAF, the biologically inactive enantiomer, and Lyso-PAF did not stimulate a significant migration of EC. This effect of PAF was not associated with a proliferative response of EC to this mediator. Moreover, the ability of PAF to stimulate the migration of EC was independent of the presence of heparin in the medium. WEB 2170, a specific PAF receptor antagonist, prevented the migration of EC induced by PAF, thus suggesting a receptor-dependent stimulation. The expression of PAF receptor gene by EC was confirmed by reverse transcriptase-PCR and Southern blot analysis. The in vivo angiogenic effect of PAF was studied in mice using a model in which Matrigel was used for the delivery of mediators. PAF induced a dose-dependent angiogenic response, which at pharmacologic concentrations (1-5 microM) did not require heparin, but at physiologic concentrations (5-50 nM) required the presence of heparin at doses that were not angiogenic per se. The angiogenesis induced by 50 nM PAF was, indeed, inhibited both by protamine and by the PAF receptor antagonist WEB 2170. The angiogenic effect of D-PAF and Lyso-PAF was not significant. Neutralizing Abs to basic fibroblast growth factor induced a slight but not statistically significant reduction of the angiogenesis induced by PAF.

Albumin and cholesterol are considered reliable outcome markers in dialysis patients; their influence, however, may also be related to non-independent factors, such as age and presence of co-morbid conditions. The aim of the study was an analysis of four outcome markers, assessed at start of dialysis: age, high risk conditions, cholesterol and albumin levels. Data were obtained from the Piedmont Dialysis and Transplantation Registry (northern Italy, about 4,400,000 inhabitants, 21 dialysis centres, open acceptance since mid-1970s, 5661 patients on file at 31 December 1992). Prevalence of albumin and cholesterol in the normal range increases with age; in each age group prevalence in the range is higher in patients at high risk. However, influence of these biochemical parameters is evident also in no-risk cohorts, thus identifying a subgroup with poorer prognosis also in the population without any identified classic risk factor. The influence of albumin, more evident in the population studied compared with cholesterol, is reflected by impaired survival of low-albumin patients (age > or = 65 high risk at 1 year: 60.7% vs 76.6%, P = 0.0052; age > or = 65 non-high risk, at 1 year: 76.5% vs 90.7%, P = 0.0001). In conclusion, albumin and cholesterol, assessed at start of dialysis, are reliable outcome markers even in elderly patients, identifying, in this high mortality cohort, a subgroup with poorer prognosis. If and how their effect may be reversed by dialysis therapy remains to be assessed.

CAPD outcomes were compared between a group of 301 diabetic patients (mean age +/- SD, 58.9 +/- 12.7 years, 55.8% males) and a group of 1689 non-diabetic patients (mean age +/- SD 57.8 +/- 14.8 years, 55.9% males) treated in 30 centres participating in the Italian Cooperative Peritoneal Dialysis Study Group from 1980 to 1989, with follow-up observation periods of 444 years (mean +/- SD, 1.48 +/- 1.24) and of 3502 years (mean +/- SD, 2.07 +/- 1.91) respectively. CAPD was the first modality for 87.2% of diabetics and 78.1% of non-diabetics (P < 0.001). The percentage of patients who needed a partner for CAPD was 45.9% in diabetics and 30.2% in non-diabetics (P < 0.001). In diabetics compared with non-diabetics, cardiovascular diseases and cachexia were nearly twice and infections other than peritonitis more than three times as frequent in causing death. In diabetics, survival was significantly worse (P < 0.0001) and the relative risk of death 2.13 times higher (P < 0.001). The technique survival and the relative risk of drop-out were not significantly different in the two groups. Clinical problems were the most important cause of drop-out among diabetics. The probability and relative risk of drop-out due to peritonitis, as well as of the first peritonitis episode, were not significantly different between the two groups and between diabetics using or not using intraperitoneal insulin. Days per patient year of hospitalization, excluding the first, were 18.4 in diabetics and 14.3 in non-diabetics. CAPD-related problems caused hospitalization in a similar way in the two groups.(ABSTRACT TRUNCATED AT 250 WORDS)

Prevalence of diabetic patients on dialysis is often considered a marker of overall acceptance rate for dialysis; however, even when acceptance policy is open, incidence of diabetic patients varies widely. Epidemiological differences of diabetes incidence all over the world partly explain the discrepancies. Incidence of diabetic patients accepted for dialysis (1981 to 82: 6 p.m.p.; 1989 to 90: 11.5 p.m.p.) differs according to age and sex in the setting analyzed (Piedmont, Northern Italian region, about 4,400,000 inhabitants, 20 dialysis centers, open acceptance since the mid-70s, yearly information on 100% of patients, gathered by a Dialysis and Transplantation Registry). Patterns changed remarkably during the 10 years considered (1981 to 90). Incidence was higher in males (10.4 p.m.p. in the period 1981 to 90), with a peak at ages 60 to 69. Incidence remained relatively stable in the younger patients, but increased in the elderly, mainly in males, rising from 6.23 in 1981 to 82 to 12.88 p.m.p. in 1989 to 90 (males, all ages). In conclusion, the demographic characteristics of diabetic patients with ESRD accepted for dialysis is changing. The stability of incidence of younger patients reassures about the open acceptance policy, at least in these ages. The increase in the elderly probably reflects the longer lifespan of diabetic patients in the overall population. The possibility of a hidden preselection must be further assessed. Future provisions of dialysis needs must take into account the trend towards an increase of this high risk, elderly population.

Resident glomerular cells have complex cytoskeletal organizations relevant to maintaining functional and structural integrity. The ability of cells to change shape, develop coordinate directed movements, replicate and interact with contiguous cells or extracellular matrix depends on cytoskeletal functions. Several cytokines directly or indirectly change cytoskeletal organization. Cytokines may alter cytoskeletal organization in cultured glomerular epithelial and mesangial cells and vascular endothelial cells mainly by mechanisms involving production of secondary mediators, such as PAF and leukotrienes that stimulate cell contraction, or PGE2 and PGI2 that stimulate cell relaxation. Consequent assembly and disassembly of microfilaments leads to functional responses of cells that may account for the physiopathological changes induced by cytokines in vivo.