Biocompatibility of a new high-permeability modified cellulose membrane for haemodialysis

The biocompatibility and solute permeability characteristicsof a high-permeability modified cellulose membrane (Hemophan-HP^®)(He-HP) were compared with those of two synthetic membranes(poly(ethylene-co-vinyl alcohol) (EVAL) and poly(acrylonitrile-co-sodiummethallyl sulphonate) (AN69)) and Cuprophan^® in a multicentre,four-way cross-over clinical trial. Cuprophan^® membranescaused significant complement activation, leukopenia, and granulocyteelas-tase release. He-HP membranes demonstrated a lesser effect,which was similar to that observed for the EVAL membrane, althoughless than that seen with the AN69 membrane. A similar orderfor the four membranes was seen for their effect on platelets.Cuprophan^® membranes provided superior small-molecule removalto the other three membranes. In contrast, Cuprophan^® wasessentially impermeable to ß₂-microglobulin, whereasHe-HP, EVAL, and AN69 allowed the removal of 60–90 mgof ß₂-microglobulin per treatment. However, a decreasein the plasma concentration of ß₂-microglobulin wasobserved only with the AN69 membrane, most probably as a resultof the ability of that membrane to adsorb proteins. Our resultsdemonstrate that high-permeability membranes of comparable biocompatibilityto some synthetic membranes can be fabricated from cellulosederivatives.     

{beta}2-Microglobulin Kinetics During Haemodialysis and Haemofiltration

Since the identification of ß₂-microglobulin as amajor component of ‘dialysis amyloid’, concern aboutits removal by different dialysis methods has been raised. Haemodialysiswith regenerated cellulose membranes increases serum ß₂-microglobulinby 10–15%. Serial measurements show a very early increaseduring cuprophan haemodialysis, the mechanism of which is asyet unknown. After cuprophan haemodialysis, serum values returnto the initial pretreatment concentrations by the time of thenext haemodialysis. In contrast to regenerated cellulose, dialysiswith polycarbonate lowers serum ß₂-microglobulin by8%, and dialysis with polysulphone by 53%. As opposed to cuprophan,after polysulphone haemodialysis the serum concentrations havenot returned to the initial pretreatment levels within 48 h.Comparison of ß₂-microglobulin removal using the samepolysulphone membrane for haemodialysis and haemofiltrationshows that ß₂-microglobulin is more effectively removedby convection than by diffusion when both treatment modes arematched for blood flow and urea clearance. Therefore, in contrast to haemodialysis with regenerated cellulosemembranes, where a transient, intradialytic release of ß₂-microglobulinis induced, significant removal is observed using, higher permeablemembranes. These findings may have implications for the generationof ‘dialysis amyloid’.     

Dialysis Arthropathy, {beta}2-Microglobulin and the Effect of Dialyser Membrane

Patients with dialysis arthropathy had the greatest mean serumß₂-microglobulin (59.5 mg/l) but there was no thresholdconcentration of ß₂-microglobulin above which allpatients developed dialysis arthropathy. Haemodialysis patientswithout dialysis arthropathy and patients on continuous ambulatoryperitoneal dialysis (CAPD) also had grossly elevated valuesof ß₂-microglobulin (47.9 mg/l and 30.7 mg/l respectively).There was a significant positive correlation between durationof treatment and serum ß₂ for the patients treatedby haemodialysis, but this was not the case for patients onCAPD. There was a significant negative correlation between residualurinary volume and serum ß₂-microglobulin for thepatients on haemodialysis without dialysis arthropathy, andalso for patients on CAPD. This was not true for the patientswith dialysis arthropathy. Both duration of treatment and residualurine volume correlated with serum ß₂-microglobulin,and therefore an analysis of covariance was used to take accountof this in comparing the groups. This showed that there wasno difference between serum ß₂-microglobulin in haemodialysispatients with and without dialysis arthropathy. However, CAPDpatients had a significantly lower corrected mean serum ß₂-microglobulinHaemodialysis with cuprophane membranes was associated withan increase in ß₂-microglobulin of 11.5%, whereashaemodialysis with polycarbonate was associated with a decreaseof 6.8% at 6 h. Our results provide circumstantial evidencethat repeated haemodialyses with cupro phane membranes may predisposelong-term haemo dialysis patients to dialysis arthropathy. CAPDpatients have lower ß₂-microglobulin concentrationsand may be less likely to develop dialysis a Long-term prospectivestudies are needed to confirm these assertions.     

Influence of Haemodialysis Membranes on {beta}2-Microlobulin Kinetics:In Vivo and In Vitro Studies

Amyloidosis of the ß₂-microglobulin (ß₂M)type is a recently recognised complication of dialysis. We studiedplasma and ultrafiltrate ß₂M in 36 chronic haemodialysispatients. Long-term dialysis with standard cuprophan membranein non-oliguric patients and with the AN69 polyacrylonitrilemembrane in oliguric patients resulted in lower plasma ß₂Mconcentrations (means±SD, 24.4±6.6 and 33.0±8.2µg/ml, respectively) than with cuprophan in oliguric patients(47.0±13.2 µg/ml, P<0.01). In acute studies, plasma ß₂M (corrected for haemoconcentration)increased, although not significantly, during cuprophan dialysis(n=10) from 40.6±12.2 to 44.8±7.6 µg/mland decreased during AN69 dialyis (n=10) from 39.4±18.0to 24.3±7.1 µg/ml (P<0.02). After 15 min ultrafiltration,ß₂M sieving coefficient in vivo was 0.33 for AN69but near zero for cuprophan. Total mass transfer of ß₂Macross the AN69 membrane during a 4-h dialysis was 142 ±45.8 mg(n=7). AN69, but not cuprophan, membrane fragments incubatedin vitro with normal or uraemic plasma exhibited avid [125-I]ß₂M binding (respectively 12.9% vs 0.47% and 6.09%vs 0.06% of total ß₂M bound at 30 mim, n=6, P<0.001for both). No in vitro generation of ß₂M from wholenormal or uraemic blood could be demonstrated during incubationwith either membrane. In conclusion, at least in some patients, standard cuprophandialysis promotes retention and possibly tissue release of ß₂Min vivo, whereas AN69 dialysis leads to ß₂M removaldue to both transmembrane transfer and, to a lesser extent,membrane binding.     

Serum {beta}2 Microglobulin and Extracellular Fluid Volume During Haemodialysis

Conflicting results have been published concerning serum ß₂microglobulin (ß₂-M) kinetics during dialysis witha cuprophane membrane which is not permeable for the protein.We have investigated the hypothesis that the apparent increaseof free serum ß₂-M could result from extracellularfluid volume (ECV) contraction. Using inulin, ECV was measuredbefore and 1 h after a dialysis session with a sodium dialysateconcentration of 145 mmol/l. Dialysis was performed either witha cuprophane or a high-flux membrane. Transcellular water shiftand changes in ß₂-M concentration were calculatedfrom total body water changes (ultrafiltration) and ECV. ECV decreased from a predialysis value of 16.6±3.51 (mean±SD)(24.9% bodyweight) to a postdialysis value of 11.9±2.11(19.8% bodyweight). Ultrafiltration was only 3.1±1.01,indicating concomitant water shift from ECV to intracellularfluid space. A significant decrease in corrected ß₂-Mconcentration was found for high-flux membranes. However, postdialysisß₂-M did not change significantly after dialysis withcellulosic membranes. In conclusion, the apparent increase of serum ß₂-Mconcentration measured during dialysis with cellulosic membranesmay be explained by ECV contraction. These results have to betaken into account for any pathogenic mechanism of the ß₂-M-associatedamyloidosis occurring in long-term haemodialysed patients.     

Determinants of Plasma {beta}2 Microglobulin Concentration: Possible Relation to Membrane Biocompatibility

Beta₂ microglobulin (ß₂m) concentrations were measuredby radioimmunoassay in the serum of haemodialysed patients.ß₂m was higher in males (n=48) than in females (n=26),i.e. 40.3±10.1 mg/l (SD) vs 31.2±8.0, P<0.01).ß₂m was not significantly higher in patients withbone cysts (37.7±11.4 mg/l vs 37.0±10.0), butmedian duration of dialysis was significantly (P<0.01) longerin patients with bone cysts (90 vs 57 months). ß₂mwas lower in patients maintained on dialysis for less than 1year and whose residual urine volume was greater than 0.1 litreper day. During one single session of dialysis, using cuprophanemembranes, ß₂m increased acutely at 15 min and hadrisen by 32.4% at the end of the dialysis session, more thancould be explained by haemoconcentration. In contrast, ß₂macutely decreased by 38.7% during a single session using polysulphonemembranes and the steady state predialysis values were lowerby 37.1% after two weeks intermittent haemodialysis with polysulphonemembranes. After re-exposure to cuprophane serum ß₂mincreased to the original value. It is concluded that ß₂m concentrations on dialysisare a function of residual urinary volume, sex, and type ofmembrane used. Data are consistent with effective removal ofß₂m by membranes with high cut-off.     

Serum {alpha}2-macroglobulin in haemodialysis patients: baseline and kinetic studies

The pathogenesis of dialysis related amyloidosis remains unresolveddespite the identification of ß₂-microglobulin (ß₂M)as the major protein constituent, as well as other proteinsbeing present in the deposits. Among the latter we have assessedthe serum concentrations of ₂-macroglobulin (₂M) both in thebaseline stage and during the haemodialysis (HD) procedure.We have also assessed the influence of the membrane on ₂M kinetics. Fifteen HD patients with histologically proven dialysis-relatedamyloidosis (DRA group) and 15 HD patients clinically and radiologicallyconsidered dialysis-related amyloidosis free (control group)were included in the baseline study. Blood was sampled the daybefore the second dialysis of the week and ₂M, ß₂Mand ₁, antitrypsin were determined along with the routine biologicalanalysis of these patients. Serum ₂M was greater in dialysis-relatedamyloidosis than in control patients (t = 2.35; P<0.026).Serum ß₂M was similar in both groups. The serum ₂Mand ß₂M correlated in patients with dialysis-relatedamyloidosis (r = 0.64; P<0.01), while no correlation wasfound in controls (r = 0.17; NS). Stepwise analysis taking thepresence of dialysis-related amyloidosis as the dependent variableretained the serum ₂M concentration as the first variable inthe model (F = 4.4; partial r = 0.38; P<0.046). The sameproteins were determined in another group of seven patients,before and hourly during HD as well as 2 and 8 h after the endof HD during nine consecutive dialyses (3 cycles of 3 HD eachusing AN69 and cuprophane membranes in a crossover design).Serum ₂M significantly increased from hour 3 and continued toincrease 2 hours post-HD (+11% and +9% with AN69 and cuprophanerespectively; P<0.001). Total proteins peaked at hour 4 (+4% and +3% P<0.01) and decreased after HD. Serum ß₂Msignificantly decreased with AN69 HD ( – 29% P<0.001)and remained unchanged during cuprophane HD. In conclusion, significant increases in serum ₂M are observedimmediately after and during the early post-dialysis periods,regardless of the membrane used. Further, serum ₂M correlateswith ß₂M only in patients with dialysis-related amyloidosis,and this variable was retained in the multivariate regressionanalysis to predict dialysis-related amyloidosis. Although thebaseline results require confirmation with larger studies, wepostulate that the present results are of relevance for dialysis-relatedamyloidosis pathogenesis since ₂M, previously identified indialysis related amyloid deposits, is closely related to acute-phasereactant proteins, and interacts with the main infiltratingcells of the deposits (macrophages). ₂M modifications couldrepresent a new manifestation of the inflammatory response tothe haemodialysis procedure.     

High-Flux Synthetic Versus Cellulosic Membranes for {beta}2-Microglobulin Removal During Hemodialysis, Hemodiafiltration and Hemoflitration

Efficient removal of ß₂ microglobulin (ß₂-M)in end-stage renal failure patients is a continuing preoccupation,as the incidence and severity of dialysis-associated amyloidosisare increasing. To evaluate comparative ß₂-M removalwe studied six stable end-stage renal failure patients duringhigh-flux 3-h haemodialysis, haemodiafiltration, and haemofiltration,using acrylonitrile, cellulose triacetate, polyamide and polysulphonecapillary devices. The reduction of plasma ß₂-M, totalremoval in ultrafiltrate/dialysate, and ß₂-M sievingcoefficients were measured by RIA. The results suggest thatconvection plays the major role in ß₂-M removal whenhigh-flux synthetic membranes are used in combination with highblood flow rates. In contrast, using the cellulose triacetatemembrane under investigation, ß₂-M removal is diminishedwhen ultrafiltration rates are increased. Accordingly, in anyfuture prospective study on the role of ß₂-M retentionin the amyloidogenesis, it is recommended that high-flux syntheticmembranes be employed rather than the type of high-flux cellulosicmembranes used in this study. The modality with which thesesynthetic membranes are used is probably less important, aslong as maximum convective transport rates are obtained. Underpresent conditions, this will imply haemofiltration or haemodiafiltrationrather than haemodialysis.     

Tissue Uptake of 125I-{beta}2-Microglobulin({beta}2-M) in Anephric Animals in the Presence or Absence of Aluminium Intoxication

In long-term haemodialysis patients a new type of amyloidosiscomposed of ß₂-microglobulin (ß₂-M) hasrecently been described. The amyloid deposition has a particularpredilection for articular structures. In the pathogenesis ofthis complication markedly elevated plasma ß₂-M concentrations,such as those observed in anuric patients, have a role. However,other as yet ill-defined factors must also be implicated, possiblecandidates being aluminium intoxication and the widely usedregenerated cellulose (cuprophan) membrane. In the present experimentalstudy, we examined tissue distribution of exogenous ß₂-Mafter i.v. injection of ¹²⁵I-ß₂-M to bilaterally nephrectomisedrats. One hundred and twenty minutes after injection, most radioactivityremained in the vascular compartment. The accumulation in tissueswas weak, and no predilection for a particular tissue becameapparent. Interestingly, chronically aluminium-overloaded, acutelyanephric rats accumulated a significantly greater amount of¹²⁵I-ß₂-M in their spleens than anephric rats withoutprior aluminium intoxication. We then attempted to induce ß₂-M amyloid depositionin rats and mice, some of whom had undergone chronic aluminiumintoxication and subcutaneous implantation of regenerated cellulosefragments for various periods of time. They were subsequentlymade anephric to obtain high plasma ß₂-M concentrations.None of the animals developed ß₂-M amyloidosis inspleen, liver, skin and mechanically altered joint synovium. In conclusion, chronic aluminium intoxication enhances splenicaccumulation of exogenous ¹²⁵I-ß₂-M in anephric rats.The factors required to form ß₂-M-amyloidosis in vivohave still to be defined.     

Highly sulphated glycosaminoglycans in articular cartilage and other tissues containing {beta}2 microglobulin dialysis amyloid deposits

BACKGROUND: Highly sulphated glycosaminoglycans (GAGs) are a common constituentof amyloid deposits and an integral component of articular connectivetissues where ß₂-microglobulin (ß₂M) amyloidis most often found. METHODS: Using alcian blue, magnesium chloride, critical electrolyteconcentration, mucin histochemistry, and immunohistochemistry,the GAGs composition of ß₂M amyloid deposits in jointcapsule and cartilage, carpal, and heart tissues of 22 uraemicpatients was determined. RESULTS: Highly sulphated GAGs were found in ß₂M amyloid depositsnot only within cartilage, where such GAGs are normally foundin high concentration, but also in other articular and extra-articularconnective tissues. Keratan sulphate was often specificallylocalized to ß₂M amyloid deposits in articular cartilageand to a lesser extent in periarticular tissues, with one caseshowing colocalization with systemic vascular amyloid deposition.Other sulphated GAGs, chondroitin 4 and 6 sulphate, dermatansulphate, and heparan sulphate were also identified in tissuescontaining ß₂M amyloid deposits, but with the exceptionof heparan sulphate (identified by mucin histochemistry) werenot specifically localized to the deposits themselves. CONCLUSION: These findings suggest that qualitative or quantitative changesin the composition of highly sulphated GAGs may play a rolein localization of ß₂M amyloid deposits in articularand extra-articular tissues.     

Mid-dilution on-line haemodiafiltration in a standard dialyser configuration

Background. Mid-dilution haemodiafiltration (HDF) results inan improved middle molecule removal compared with standard HDF.The OLpr^TM MD 190 haemodiafilter represents a new dialyser designexclusively for mid-dilution on-line HDF. Compared with standardhaemodialysers, structural changes in the headers allow theinfusion of high replacement fluid volumes after a first post-dilutionand before a second pre-dilution stage. Methods. We compared in vitro the new device [blood flow (Q_B)400 ml/min, substitution flow (Q_S) 100 and 200 ml/min, dialysateflow (Q_D) 800 ml/min] with a conventional high-flux dialyserof the same surface area in haemodialysis (HD) (Q_D 500 ml/min)and post-dilution HDF (at Q_S 60, Q_D = 500 ml/min and at Q_S 100,Q_D = 800 ml/min) modes. Subsequently, we performed an initialclinical application of the new device in six mid-dilution HDFtreatments of five end-stage renal disease patients (Q_B 400ml/min, Q_S 200 ml/min, Q_D 800 ml/min, treatment duration 205±23min). Results. In vitro urea and ß₂-microglobulin clearancesin mid-dilution HDF were, respectively, 309.2±5.5 and144.4±15.2 ml/min (Q_S 100) and 321.6±4.1 and 204.9±4.1ml/min (Q_S 200), compared with 278.6± 17.2 and 94.0±7.6ml/min in HD, and 310.8±10.2 and 123.0±6.5 ml/min(Q_S 60) and 323.6±11.2 and 158.0±10.3 ml/min (Q_S100) in post-dilution HDF. The in vivo trials showed the clinicalutility of the device and confirmed the in vitro data: ureaand ß₂-microglobulin clearances were, respectively,324.6± 10.9 and 207.9±29.3 ml/min, while reductionratios were 75.0±5.5 and 83.6±4.7%. Conclusion. Our preliminary results need confirmation in a prospectivecross-over study. However, the Nephros MD 190 haemodiafilterpromises to be a true technological step ahead in terms of improvedß₂-microglobulin removal.     

Continuous Ambulatory Peritoneal Dialysis vs Haemodialysis: A Lesser Risk of Amyloidosis?

We compared plasma beta-2-microglobulin ß₂M at a 1-yearinterval in 25 CAPD patients and 25 patients haemodialysed withcuprophane membranes and matched for residual renal functionand duration of renal replacement therapy. Plasma ß₂Mremained lower in CAPD patients throughout the study, and increasedsignificantly with time both in CAPD and haemodialysis patients,as renal function decreased. In both groups, plasma ß₂Mwas negatively correlated with residual creatinine clearance,the influence of the latter being much greater in haemodialysis,as demonstrated by comparison of the regression lines. In haemodialysis,but not in CAPD. plasma ß₂M also correlated with timeon dialysis. In CAPD patients. the daily peritoneal output averaged 38 mg(range 16–59 mg), and was directly correlated with plasmaß₂M CAPD thus allows a significant peritoneal removal of ß₂Mwhich progressively takes over from the declining renal function,resulting in lower plasma ß₂M than in matched haemodialysispatients. However, the peritoneal removal of ß₂M remainsinsufficient and values increase with time as renal functiondeclines. Thus, if ß₂M amyloidosis is related to raisedplasma levels, the risk of ß₂M amyloidosis in CAPDshould simply be delayed as compared to haemodialysis.     

Beta2 adrenergic antagonist inhibits cerebral cortical oxygen delivery after severe haemodilution in rats

Background. Haemodilution has been associated with neurologicalmorbidity in surgical patients. This study tests the hypothesisthat inhibition of cerebral vasodilatation by systemic ß₂adrenergic blockade would impair cerebral oxygen delivery leadingto tissue hypoxia in severely haemodiluted rats. Methods. Under general anaesthesia, cerebral tissue probes wereplaced to measure temperature, regional cerebral blood flow(rCBF) and tissue oxygen tension (P_BrO₂) in the parietal cerebralcortex or hippocampus. Baseline measurements were establishedbefore and after systemic administration of either a ß₂antagonist (10 mg kg^–1 i.v., ICI 118, 551) or saline vehicle.Acute haemodilution was then performed by simultaneously exchanging50% of the estimated blood volume (30 ml kg^–1) with pentastarch.Arterial blood gases (ABGs), haemoglobin concentration (co-oximetry),mean arterial blood pressure (MAP) and heart rate (HR) werealso measured. Data were analysed using a two-way ANOVA andpost hoc Tukey's test [mean (SD)]. Results. Haemodilution reduced the haemoglobin concentrationcomparably in all groups [71 (9) g litre^–1]. There wereno differences in ABGs, co-oximetry, HR and MAP measurementsbetween control and ß₂ blocked rats, either beforeor 60 min after drug or vehicle administration. In rats treatedwith the ß₂ antagonist there was a significant reductionin parietal cerebral cortical temperature, regional blood flowand tissue oxygen tension, relative to control rats, 60 minafter haemodilution (P<0.05 for each). These differenceswere not observed when probes were placed in the hippocampus. Conclusion. Systemic ß₂ adrenergic blockade inhibitedthe compensatory increase in parietal cerebral cortical oxygendelivery after haemodilution thereby reducing cerebral corticaltissue oxygen tension.     

Enhanced Gastrointestinal Absorption of Aluminium in Uraemia: Time Course and Effect of Vitamin D

The present study examines the time course of aluminium absorptionin uraemic rats vs controls and investigates the effect of vitaminD. Following an oral load of 410 µmol aluminium there wasa significant increase in the urinary excretion rate of aluminiumas early as 60 min in uraemic rats. Compared with controls thisincrease was significantly greater in uraemic animals and maximumexcretion rates (77±49 vs pre-load 2±1 nmol Al/h)were achieved after 2 h. When vitamin-D-deficient rats with normal renal function werecompared with vitamin-D-replete controls, the latter excreteda significantly greater amount of the oral dose of aluminiumin their urine (727±361 vs 359±l40nmol Al/5d;P<0.02) and the post-load increase in the serum aluminiumconcentration was more pronounced in the vitamin-D-replete animals. Aluminium administered i.v. resulted in similar urinary aluminiumexcretion rates in both groups. In uraemicrats, however, regardlessof their vitamin D status, adminis tration of 1,25(OH)₂D₃ hadno effect on the amount of urinary aluminium excretion afteroral or i.v. loads. These findings suggest that although in rats with normal renalfunction aluminium absorption appears to be partly vitamin Ddependent, 1,25(OH)₂D₃ does not further augment the enhancedgastrointestinal absorption of aluminium in uraemia.     

Clinical comparison of high-flux cellulose acetate and synthetic membranes

Solute transport and alterations in complement and clottinginduced by a new high-flux cellulose acetate membrane (CA-HF800-E,Diaphan^®) were compared with those for cellulose triacetate(CTA) and polysulphone in a cross-over clinical study The membranes are similar in their small-molecule removal Serumß₂-microglobulin decreased with all membranes butthe decrease was independent of membrane type. Associated withß₂-microglobulin removal was a protein loss whichaveraged 2636 mg for Diaphan^®, 4937 mg for CTA, and 2500mg for polysulphone. Albumin presence in the dialysate was lessthan the limit of detection (5mg/l) but for each of the membranes,occasional readings above the limit of detection were noted. C3a generation for Diaphan^® is comparable with that forCTA and polysulphone, but differed for C5a and neutropenia.A highly significant correlation of the area under the concentrationtime curve of the two complement components was noted for thecellulose based membranes (r = 0.875, P = 0.0002 for Diaphan^®,r = 0.823, P = 0.006 for CTA) this relationship was less markedfor polysulphone (r = 0.396, P = 0.29) Induction of clotting characterized by the thrombin-antithrombinIII (TAT) complex were similar for the three membranes, as werechanges in platelet counts. Our findings indicate that while it is possible to modify celluloseto produce a membrane whose solute transport and biocompatibilityis similar to synthetic membranes such as polysulphone, thestructural modifications induce considerable differences inthe amount of protein lost.     

Ultrasonographic detection of thickened joint capsules and tendons as marker of dialysis-related amyloidosis: a cross-sectional and longitudinal study

Dialysis-related amyloidosis is characterized by a ß₂-microglobulin(ß₂M) infiltration of joint synovia, tendons and capsules.We report a cross-sectional ultrasonographic evaluation of supraspinatustendon and femoral neck capsule thickness in 49 patients onlong-term haemodialysis. Ultrasonographic evaluation was repeated21±4 (SD) months later in 16 patients. Normal valuesfor the supraspinatus tendon and femoral neck capsule were definedin a group of control subjects without history or signs of jointdisease. Among the 49 patients, aged 21–86 (median 59) years, dialysedfor 1–228 (median 97) months, 33 had at least one abnormaljoint. The prevalence of patients with at least one and at leasttwo abnormal joints, the number of abnormal joints per patient,and the thickness of the supraspinatus tendon and femoral neckcapsule increased significantly with dialysis duration (P<0.001 for all parameters considered). By multiple linear regressionanalysis, mean thickness of the supraspinatus tendon was positivelyrelated to both dialysis duration (P< 0.0001) and age (P= 0.036) independently. All (n=11) patients with radiological and/or histological evidenceof dialysis-related amyloidosis at the time of ultrasonographyhad thickened supraspinatus tendon and/or femoral neck capsule;which were also thickened in an additional 22 patients withoutradiological evidence of dialysis-related amyloidosis. Threedied within 5–10 months of the ultrasonographic investigation: post-mortem examination of the periarticular tissue confirmedthat the detected thickening was due in all three to ß₂Mamyloid infiltration. Sixteen patients underwent a second ultrasonographic evaluation21±4 months later. In nine patients on dialysis for <60months at the time of the first evaluation, mean femoral neckcapsule thickness increased significantly (7.0±0.8 to8.2±2.3mm, P = 0.017) whereas mean supraspinatus tendonthickness increment was not significant (6.6±0.4 to 7±0.8mm, P=0.23). In the seven other subjects dialysed for <60months, neither the supraspinatus tendon nor femoral neck capsulethickness changed. We suggest that ultrasonographic measurement of supraspinatustendon and femoral neck capsule thickness is a useful, non-invasivetool to detect and monitor dialysis-related amyloidosis.     

Effect of autonomic neuropathy on ventilatory response to progressive hypercapnia in dialysis patients

The impact of autonomic neuropathy (common in patients on haemodialysis)on ventilatory response to hypercapnia has been studied. We investigated cardiac reflex tests in 20 patients on chronichaemodialysis (8 patients were found with and 12 without neuropathyof the autonomic nervous system). Using the hyperoxic CO₂-rebreathingmethod (according to Read), we tested the above-mentioned twogroups of patients and compared them with 14 healthy controlsubjects. Accumulation of CO₂ in blood with hyperoxic CO₂ rebreathingstimulates central chemoreceptors, and therefore causes a progressiverise in minute ventilation. In patients with autonomic neuropathy (n=8), ventilatory responseto increasing pCO₂ was significantly lower than that in thecontrols (1.7±0.3 versus 3.2±0.5 l/min/mmHg, P<0.001).On the other hand ventilatory response in patients without autonomicdamage (n=12) showed no significant difference when comparedto controls (3.1±0.8 l/min/mmHg). There were no differencesin lung function, arterial blood gas analysis, blood chemistry,duration on dialysis, and demographic data when comparing thepatients with and those without autonomic damage. Our analysis shows different patterns of ventilatory responseto increasing pCO₂ in patients on haemodialysis. Autonomic neuropathyhas to be considered when rebreathing tests are interpreted.The clinical relevance of these findings needs further investigation.     

Calcium carbonate (CaCO3): an efficient and safe phosphate binder in haemodialysis patients? A 3-year study

In an uncontrolled study, 22 dialysis patients (46±14years, duration of dialysis 20±11 months) were treatedwith CaCO₃ over a period of up to 3 years to lower their serumphosphate. The use of 4.5–9 g CaCO₃ daily over a periodof 9 months led to a reduction of mean serum phosphate from2.51 to 1.51 mmol/1 in 77% of patients, with a simultaneousincrease in mean calcium concentration from 2.23 to 2.47 mmol/1,and an improved control of secondary hyperparathyroidism byreduction in mPTH from 1552 to 1032 pg/ml and in APH activityfrom 6.25 to 4.55 µmol/s/1. In long-term CaCO₃ treatment of up to 3 years, however, a constanteffective phosphate reduction could not be achieved. There wasa progression (77%) of pre-existing microcalcification and anew appearance (42%) of microcalcification in vessels and soft-tissueareas of the hand. The percentage of patients with soft-tissuecalcification increased from 43 to 67% during a treatment periodof 3 years. We conclude that CaCO₃ alone is not suitable on a long-termbasis for phosphate reduction in dialysis patients.     

Prevalence of dialysis-related amyloidosis in diabetic patients

It has recently been shown that ß₂-microglobulin isolatedfrom amyloid deposits in dialysis patients is modified by advancedglycation (AGE). In this context it appeared of interest toexamine in a cross-sectional multicentre study whether dialysis-relatedamyloidosis, as evaluated by X-ray assessment of cysts in themetacarpal bones, was different in diabetic patients on maintenancehaemodialysis for more than 5 years time compared with matchednon-diabetic controls. We evaluated the hand skeleton of 75diabetic patients (9 type I, 66 type II; 35 male, 40 female;median age 64 years, range 31–86; median duration of dialysis7 years, range 5–17). They were compared with 150 patientswithout diabetes mellitus who were matched for age, gender andduration of dialysis. Hand X-rays were centrally evaluated byone radiologist unaware of the underlying clinical diagnosis.The overall frequency of amyloid cysts was 9/75 (12%) in diabeticpatients (95% confidence interval 4.6–19.3%) and 28/150(19%) in matched controls (95% confidence interval 12.4–24.9%).The results indicate that diabetes mellitus does not conferan increased risk of dialysis-related amyloid cysts. The resultsare of interest with respect to the mechanism of amyloid formation.     

Adherence of human monocytes to haemodialysis membranes

In the present study we evaluated spontaneous and stimulatedadherence of human monwytes to regenerated cellulose and polyacrylonitrile(AN69) membranes. Spontaneous adherence at 60 min was significantlyhigher for regenerated cellulose (28±2%, P<0.001)than for AN69 (11±2) membranes. Stimuli such as bacteriallipopolysaccharide, TNF, interleukin-1 and -6 as well as platelet-activatingfactor, but not IL4, significantly enhanced adherence at 60min to AN69 (28 to 30%). In contrast, adherence was not furtherinducible in the presence of regenerated cellulose. Both spontaneousand cytokine/bacterial lipopolysaccharide-stimulated adherencewere significantly reduced by SDZ-63072, a specific plateletactivatingfactor receptor antagonist. This difference in sensitivity ofmonocyte adherence reflects probably the intrinsic ability ofregenerated cellulose to provide maximal spontaneous monocyteadhesion. These data suggest that PAF may act as an adherencemediator. This is in line with the ability of regenerated celluloseto directly stimulate monocytes to synthesize plateletactivatingfactor and with the ability of cytokines and bacterial lipopolysaccharideto stimulate its synthesis. Although AN69 has a low adherencepotential, bacterial lipopolysaccharide or cytokines may bluntthe biocompatibility of this membrane.