peritoneal dialysis - core

Kidney International, Vol. 45 (1994), pp. 1226—1240

NEPHROLOGY FORUM

Peritoneal dialysisPrincipal discussant: CIDI0 CHAIM0vITz

Soroka Medical Center of Kupat Holim, Ben Gurion University of the Negev Center for Health Sciences, Beer-S heva, Israel

A 46-year-old man with end-stage renal failure due to polycystickidney disease, who had been treated by continuous ambulatoryperitoneal dialysis (CAPD) for 33 months, was admitted to the SorokaMedical Center because of 2 hours of fever, abdominal pain, nausea,and cloudy peritoneal fluid. On physical examination he appeareduncomfortable and complained of severe abdominal pain. His temper-ature was 39°C; pulse, 108 beats/mm; and respirations, 24 breaths/mm.The blood pressure was 150/95 mm Hg. The lungs were clear; the heartrhythm was regular, and an apical grade-2 holosystolic murmur wasaudible. Abdominal examination disclosed diminished bowel soundsand diffuse abdominal tenderness. The exit site of the Tenckhoffcatheter was normal. The lower extremities were normal and peripheralpulses were intact. The peritoneal effluent was very cloudy and con-tained 7000 cells/mm3, of which 95% were polymorphonuclear leuko-cytes. The patient was treated initially with a combination of intraperi-toneal vancomycin and aminoglycoside. Dialysate cultures grewStaphylococcus aureus and rifampin was added to the treatment. Theperitoneal dialysate gradually cleared, but the patient continued to havefever and severe abdominal pain for 5 days, and the dialysate culturebecame negative only after 7 days of treatment.

Until his latest admission, the patient's fluid balance, nutrition, andbiochemical values had been well controlled. He had been treated withfour peritoneal exchanges daily; three changes were performed withperitoneal dialysis solutions containing 1.5% glucose and the fourthwith 2.5%. The calcium concentration of the dialysis solution was 2.5mEq/liter. He received 600 mg calcium carbonate 3 times/day during

This Forum was presented at the 12th International Congress ofNephrology, Jerusalem, Israel, June, 1993. Presentation of this Forumis made possible by grants from Merck Sharp & Dohme International;Dialysis Clinic, Incorporated; Marion Merrell Dow, Incorporated; andMead-Johnson Pharmaceuticals.

© 1994 by the International Society of Nephrology

meals, 0.25 pg/day of vitamin D3, and subcutaneously administerederythropoietin, 4000 U/week. The patient's weekly combined creatinineclearance (renal and peritoneal clearance) was 60 liters. Laboratorytests one month before the episode of peritonitis showed a BUN of 64mg/dl; serum creatinine, 9.2 mg/dl; uric acid, 8.7 mg/dl; sodium, 138mEq/liter; potassium, 4.3 mEq/liter; calcium, 10.3 mg/dl; phosphorus,5.3 mg/dl; albumin, 3.8 g/dl; globulin, 2.7 g/dl; alkaline phosphatase, 110U/liter (normal, 1—125 U/liter); and cholesterol, 260 mg/dl. The plasmaintact PTH was 110 pg/mI (normal, 20—55 pg/ml). The hematocnt was30%. A peritoneal equilibration test (PET) 4 months prior to theperitonitis, performed 4 hours after the infusion of peritoneal dialysissolution containing 2.5 g/dl glucose, showed a drain volume of 2550 ml;the ratio of the dialysate glucose concentration at 4 hours dwell time tothe dialysate glucose concentration at zero dwell time (D/DO) was 0.48[1].

In the months following his hospitalization for peritonitis, symptomsof fluid retention appeared; systemic hypertension developed, and theplasma urea and creatinine levels rose. A repeat PET showed a severeultrafiltration defect (a drain volume of 2080 ml) and a DIDO glucoseratio of 0.42 [1]. Other laboratory findings included a BUN of 108 mg/dl;creatinine, 14.6 mg/dl; calcium, 10.2 mg/dl; phosphorus, 6.8 mg/dl;albumin, 3.2 g/dl; globulin, 2.7 g/dl; and hematocrit, 30%. The CAPDwas stopped and hemodialysis treatment was initiated. When theTenckhoff catheter was removed, the surgeons found a marked thick-ening of the parietal peritoneum.

Discussion

DR. CIDL0 CHAIM0vITz (Head, Department of Nephrology,Soroka Medical Center of Kupat Holim, and Professor ofMedicine, The Ben Gurion University of the Negev Center forHealth Sciences, Beer-Sheva, Israel): This patient had beensuccessfully treated with CAPD for about 3 years. His perito-neal creatinine clearance was normal, and accordingly his fluidbalance and nutritional status were excellent. Thirty-threemonths after commencement of CAPD, he had severe perito-nitis due to Staphylococcus aureus. In the following months,signs of under-dialysis and peritoneal ultrafiltration failurerequired his transfer to hemodialysis. Clinical data, radiographsof the peritoneal cavity, and the surgeon's description of thepentoneum led to a diagnosis of peritoneal fibrosis.

Although rare, a decrease in ultrafiltration is a worryingcondition in patients on CAPD [2—151. In most cases thedecrease is due to an increased permeability of the peritoneumto glucose, which is rapidly absorbed from the dialysate.Consequently the dialysate/serum osmotic gradient dissipatesquickly and retards the ultrafiltration [1, 3, 16]. Patients withthis condition, known as type-I ultrafiltration (UF) failure, havedifficulty in removing fluid during dialysis, although the ureaand creatinine clearances are maintained or even increased.This impairment in ultrafiltration occurs more frequently inpatients on long-term peritoneal dialysis, although it can be atemporary condition lasting only a few weeks after acute

1226

EditorsJORDAN J. COHENJOHN T. HARRINGTONNIcoLAos E. MADLAS

Managing EditorCHERYL J. ZUSMAN

State University of New York at Stony Brookand

Tufts University School of Medicine

Case presentation

Nephrology Forum: Peritoneal dialysis 1227

peritonitis. In a number of patients with type-I UF failure whounderwent peritoneal biopsies, the mesothelial cell layer wasseverely damaged [7, 17]. The condition can be treated by usinghypertonic dialysis solution more frequently and/or by shorten-ing the length of dwell time. Less often, the impaired ultrafil-tration is due to a reduction in the surface area of the peritonealmembrane (type-I! ultrafiltration failure) and is associated witha normal or low peritoneal permeability to glucose [3, 6, 7, 16].Type-Il UF failure occurs most frequently in patients recover-ing from severe and prolonged peritonitis, in which the patho-gen is usually Staphylococcus aureus or fungus. Peritonealbiopsies show that the delicate pentoneal structure is replacedby extensive scar tissue, and laparotomies have revealed mul-tiple pentoneal adhesions [3, 15]. In the patient under discus-sion, the peritoneal damage can be classified as type-Il ultrafil-tration failure. As happened in the patient described here, mostpatients in this category must change their treatment fromCAPD to hemodialysis.

In this Forum, I plan to review a variety of documentedinvestigations published to date that address some of the majorproblems arising in CAPD treatment.

Pathogenesis of peritoneal fibrosis

Pentoneal biopsies from uninfected CAPD patients revealsignificant alterations in the structure of the peritoneal mem-brane [18—20]. Large numbers of mesotheial cells were seen inthe process of detaching from the peritoneal surface and beingreplaced by a lively upward migration of precursor cells.Damage to the mesothelial cells, evidenced by the almostcomplete disappearance of the microvilli, also has been seen inbiopsy specimens from CAPD patients.

The histologic changes might result from exposure to theperitoneal dialysis solution, which is highly unphysiologic [19,20]. The low biocompatibility of the available dialysis solutionis due to its very low pH, high osmolality, and high glucose andlactate content. This hypothesis is supported by recent studiesin vitro demonstrating that severe damage is caused by expo-sure of primary cultures of human omentum-derived mesothe-lial cells to the commercially available peritoneal dialysis solu-tion (CDS) [21, 22].

Dobbie recently discussed the pathogenesis of peritonealfibrosis in CAPD patients [23]. He presented evidence stronglysupporting the concept that severe injury to the peritonealmesothelial cells is the initiating event that ultimately leads tofibrosis of the peritoneum. If, in addition to the chronic irritanteffect of the dialysis solution, a severe or prolonged peritonitisoccurs, the mesotheial cell injury may be so extensive that thehealing process cannot keep pace, thus leaving large areas ofsubmesothelial tissue denuded of mesothelial cells and coveredby a thick mantle of fibnn in their place. This fibrin layer, if notdegraded, will progressively organize into fibrous tissue.

The submesothelial tissue also can be damaged by directcontact with the extremely high concentrations of glucose in thedialysis solution through a process of nonenzymatic glycationof the proteins [23]. Glycated collagen fibers are less solubleand more resistant to degradation than is normal collagen; thisdifference can lead to the accumulation of extracellular matrixin the submesothelial layer.

Fibrogenic cytokines and ultrafiltration failure

Growth factors in the dialysate have been speculated as apathogenic mechanism in peritoneal fibrosis and consequentultrafiltration failure [3, 24, 25]. So-called fibrogenic cytokines,such as platelet-derived growth factor (PDGF), transforminggrowth factor beta (TGF/3), interleukin-1 (IL-i), and tumornecrosis factor alpha (TNFa), can be produced by activatedmononuclear phagocytic cells. Both PDGF and TGF/3 directlystimulate fibroblast proliferation, whereas IL-i acts indirectly byinducing fibroblasts to secrete PDGF [26—28]. Transforminggrowth factor f3 stimulates the synthesis of matrix compounds,including fibronectin, collagen, and proteoglycan [29]. Shimokadoet al demonstrated that activated peritoneal macrophages ofpatients treated with CAPD can secrete a growth factor thatstimulates proliferation of fibroblasts [30]; a significant portion ofthis mitogenic activity was accounted for by a molecule similar toPDGF. This issue was further investigated by Breborowicz andcolleagues, who demonstrated that culture medium, conditionedby peritoneal leukocytes, stimulates proliferation of fibroblasts[25].

Selgas and coworkers recently described mitogens found inthe dialysate of CAPD patients that stimulated proliferation ofmouse and human fibroblast [31]. Interestingly enough, mito-genic activity of the dialysate was greater in patients with lowultrafiltration, but the factors responsible for the connective-tissue cell proliferation were not defined. Lamperi et al reporteddata supporting the view that excessive intraperitoneal synthe-sis of cytokines plays a significant role in the pathogenesis ofperitoneal fibrosis and consequent peritoneal functional failure[24]. In comparison to patients with normal ultrafiltration ortype-I ultrafiltration failure, they found a higher concentrationof IL-i and interferon-gamma (IFNy) in the dialysate of patientswith type-Il UF failure. In patients with type-Il UF failure,peritoneal lymphocytes and macrophages stimulated in vitroreleased significantly higher quantities of IL-i and IFNy ascompared with cells from patients with type-I UF failure ornormal UF. There is still much to learn regarding the role ofcytokines in peritoneal fibrosis in patients undergoing CAPD.Moreover, we need to investigate the feasibility of therapeuticagents which, by inhibiting the synthesis of collagen and theproliferation of fibroblast, might improve this condition.

Histologic evidence of fibrous tissue accumulation in thesubmesothelial layer is not always associated with clinicalevidence of peritoneal membrane dysfunction. Peritoneal fibro-sis documented by biopsies of patients undergoing CAPD hasbeen reported to occur even though the ultrafiltration functionand peritoneal clearance were normal. It is generally believedthat a very small fraction of the total peritoneal surface isactually involved in the dialytic procedure [18]. Only that areain strict juxtaposition with the peritoneal capifiaries is directlyinvolved in the peritoneal fluxes of the fluids and solutes [18].One can therefore reasonably assume that a wide deposition offibrotic tissue in the membrane will be required to causedysfunction of the peritoneal membrane.

Macrophages and peritoneal host defense

In spite of great improvements in CAPD techniques, partic-ularly in the connecting systems, peritonitis remains the mostcommon cause for patients leaving a CAPD program [32—35].

1228 Nephrology Forum: Peritoneal dialysis

Growing evidence suggests that CDS has a detrimental effect onthe host defense mechanism of the peritoneal cavity [36—381. Iwould like to review the elements involved in the host defenseagainst micro-organisms invading the peritoneal cavity and howCDS suppresses their function.

Macrophages are the predominant cells in the dialysate ofuninfected patients undergoing acute or chronic peritonealdialysis. The dialysate-elicited macrophages have functional,immunophenotypic, and cytochemical features that differuniquely from normal peritoneal macrophages [39—42]. Someinvestigators maintain that the pentoneal macrophages ofCAPD patients are immature cells, reflecting the influx ofyounger monocytes due to the rapid turnover of the pentonealcells. Goldstein et al found that peritoneal macrophages ofCAPD patients exhibited a low uptake of eicosanoid precursorsand lack the membrane-bound enzyme ecto-5' nucleotidase thatnormally accumulates in mature macrophages [39]. Using theseobservations, the authors suggested that the peritoneal macro-phages of CAPD patients constitute an immature monocyte-likepopulation caused by the daily shedding of 30 million to 40million of these cells into the dialysate. However, subsequentcharacterization of functional markers of dialysate-elicited pen-toneal macrophages and research into the pattern of endoge-nous peroxidase activity of these cells supported the notion thatthey are phagocytes stimulated by a state of chronic inflamma-tion in the peritoneal cavity. For instance, Lewis and Holmesfound that peritoneal macrophages from patients treated withCAPD, contrary to macrophages in their peripheral blood,displayed increased binding of the chemotactic peptide C5a,increased expression of Fc receptors, and increased expressionof HLA DR and CD14 antigens [42]. Cytochemical studiescarried out in rats by Bos et al demonstrated that an intraperi-toneal injection of commercial dialysis solution changes thepattern of peroxidase activity of the peritoneal cells: 24 hoursafter dialysis fluid infusion, the number of exudate macrophagesdisplaying peroxidase activity in the lysosomal granules rosesharply [43]. This cytochemical pattern was similar to thatdescribed in the dialysate-elicited peritoneal macrophages ofuninfected patients [44]. The authors suggested that the com-mercial dialysis solution used in CAPD induces a state ofchronic sterile inflammation in the pentoneal cavity, therebyactivating the macrophages [43, 45].

The functional capacity of the macrophages harvested afteran intraperitoneal infusion of CDS also has been the object ofextensive research. However, controversy continues amongscientists as to the results. Peterson et al [37] and Davies andcolleagues [40] demonstrated that the ability to mount a phago-cytic respiratory burst by stimulated peritoneal macrophageswas greater for patients treated with CAPD than for normalcontrols. Moreover, the candidicidal activity of the dialysate-elicited macrophages was tenfold higher than that occurring innormal peritoneal macrophages. On the other hand, someinvestigations have described a normal or slightly decreasedbactericidal capacity of the dialysate-elicited macrophages rel-ative to peritoneal macrophages from healthy controls [39, 46].The reason for the considerable controversy regarding thefunctional characteristics of the peritoneal macrophages inpatients treated with CAPD is not clear. Detailed clinicalstudies will be necessary to establish the relationship betweenperitoneal macrophage function and such clinical features as

nutrition, adequacy of dialysis treatment, and the sufficiency ofcalcitriol replacement. Bearing these reservations in mind, Ibelieve that the weight of evidence supports the contention thatdialysate-elicited peritoneal macrophages are normal or stimu-lated cells.

The peritoneal macrophages in patients treated with CAPDare viewed as the peritoneal cavity's first line of defense againstinvasion of microorganisms. However, after the infusion of 2 to3 liters of dialysis solution, the number of macrophages in theperitoneal cavity drops to values as low as 103—104/ml ofeffluent. This dilution makes the probability of macrophage-microorganism interaction very low in the early stages ofperitonitis [47]. Thus it would appear that the efficacy of themacrophages, which greatly exceeds innate functions such asphagocytosis and killing of invading bacteria, is mostly accom-plished by the production of potent inflammatory cytokinessuch as IL-i and TNFa. These cytokines can interact andstimulate other pentoneal resident cells and thus amplify in-flammatory signals originating in the macrophages. To illustratesuch a mechanism, I would like to demonstrate how themesothelial cells, which constitute the majority of normalperitoneal resident cells, are able to amplify inflammatorymessages from the macrophages in the early stages of peritoni-tis. This interactive process emphasizes their contributory rolein the inflammatory reaction of the peritoneum following micro-bial invasion.

Cytokine synthesis by peritoneal mesothelial cells

I infer from recent intensive investigations that the peritonealmesothelial cells have ceased to be considered as a passivelining membrane and have come to be viewed as the target formacrophage-related cytokines [48]. Because the mesothelialcells are strategically placed in close juxtaposition both toperitoneal macrophages and underlying mesothelial microvas-cular endothelium, these cells are well situated to regulate thebi-directional interaction between the macrophages and thevascular cells.

At our laboratory in Beer-Sheva, my colleagues and I re-cently investigated the ability of human pentoneal mesothelialcells in culture to produce interleukin-l when stimulated eitherby E. coli lipopolysaccharide (LPS), TNFa, or IL-la singly, orby LPS combined with any one of these cytokines [49]. Lipo-polysaccharide is the main inflammatory product of gram-negative bacteria; TNFa and IL-i are released into the perito-neal fluid when macrophages come in contact with bacteria. Wetried to simulate the inflammatory events occurring duringbacterial peritonitis. Stimulation by LPS elicited a small butsignificant release of the two forms of IL-i: IL-i a and IL- ip,but no significant stimulatory effect was seen when mesothelialcells were incubated with either TNFa or IL-ia alone. How-ever, the combination of LPS with either IL-la or TNFasynergistically increased IL-i release by the mesothelial cells.The presence of mRNA of IL-la and IL-1f3 was determined bypolymerase chain reaction (PCR) amplification of reverse tran-scribed RNA. We detected expression of IL-ia and IL-ipmRNA when the mesothelial cells were stimulated with eitherLPS, TNFa, or IL-ia , but the combination of LPS with eitherone of the cytokines had no synergistic effect. We concludedthat the synergistic effect of combinations of LPS with TNFa or


IL-la on the release of IL-i probably occurs at the post-transcriptional level.

The expression of the IL-i gene and the translation of IL-lmRNA are distinct and dissociated processes [50, 51]. Forexample, certain stimuli such as Cia or cell adherence to a glasssurface trigger IL-i gene expression without translation intoprotein. The cells containing untranslated IL-i mRNA are"primed," and even sub-stimulatory levels of other activators,such as LPS or IL-i itself, induce a translational signal. How"priming" in the mesothelial cells constitutes one of the mech-anisms whereby LPS and TNFa are made highly synergistic inthe release of IL-l requires further investigation.

I believe the results of our work highlight the important roleof the mesothelial cells in the host defense mechanism of theperitoneal cavity. First, we have demonstrated that stimulatedhuman peritoneal mesothelial cells produce a potent pleiotropiccytokine, like IL-l, capable via other IL-l-inducible cytokinesof amplifying macrophage inflammatory signals. Second, IL-i isan important modulator of the vascular endothelial cell functionand thus might interact in a paracrine fashion with the neigh-boring peritoneal capillary network. As a consequence, theendothelial cells will be stimulated into participating in theinflammatory process by synthesizing vasodilatory molecules,neutrophil chemotactic peptides, and by expressing adhesionmolecules; as a result, more inflammatory cells will be recruited[52—56].

Expression of molecules that mediate leukocyte-endothelialcell adhesion is an early prerequisite in the inflammatoryresponse [54, 55, 57]. These molecules facilitate both adhesionto endothelial surface and transendothelial migration of leuko-cytes at the site of the inflammation. After crossing the perito-neal capillary wall, the migrating leukocytes also must traversethe mesothelial cell layers to reach the inflamed site in theperitoneal cavity. Evidence reveals that adhesion moleculesalso are expressed on the surface of the mesothelial cells. Jonjicand colleagues recently found the expression of intercellularadhesion molecule-l, ICAM-l, and the vascular cell adhesionmolecule YCAM- 1 in ascites-derived mesothelial cells [58]. Theexpression of these adhesion molecules could be up-regulatedby exposure of the mesothelial cells to TNFa and IFN7. Theirdata also revealed that activation of monocytes with LPS andIFNy considerably enhanced their adherence to the mesothelialcells, a process that can be blocked by anti-CD18, the leukocyteligand of ICAM-l. In a preliminary communication, Zeille-maker et al described the presence of ICAM-l and VCAM-l,but not endothelial leukocyte adhesion molecule-l (ELAM-l),in monolayers of human peritoneal mesothelial cells [59]. Theyalso found that stimulation of the polymorphonuclear leuko-cytes by the chemoattractant C5a was necessary for significantbinding of the leukocytes to the mesothelial cells. In ourlaboratory we also demonstrated the expression of ICAM-l inhuman mesothelial cells and a strong binding of LPS-stimulatedleukocytes to the mesothelial monolayer (Fig. 1).

Further investigations have been undertaken recently explor-ing a role for the mesothelial cells in regulating inflammatoryresponses within the peritoneal cavity. Like the pleural me-sothelium, the peritoneal mesothelial cells respond to stimula-tion by IL-i and IFN7 by synthesizing the highly potentneutrophil chemotactic factor interleukin-8 (IL-8) [58]. Produc-tion of IL-8 by omentum-derived mesothelial cells also was

recently reported by Topley et al [60]. In addition to being apotent leukocyte chemoattractant, IL-8 stimulates the releaseof superoxide and degranulation by leukocytes as well asleukotriene synthesis [61]. Although peritoneal macrophagesappear to be the predominant source of IL-8, the close prox-imity of the mesothelial cells to the peritoneal capillaries makesthem uniquely situated to regulate the recruitment of neutro-phils in the peritoneal cavity, at least in the early stages of theinfection.

Human peritoneal mesothelial cells also can be induced tosynthesize interleukin-6 (IL-6), a polypeptide with a complexspectrum of biologic effects [58, 62]. Interleukin-6 activity

A

B

Fig. 1. Adherence of unstimulated (A) or LPS-stimulated (B) polymor-phonuclear leukocytes to omentum-derived human mesothelial cellmonolayer. Polymorphonuclear leukocytes were stimulated for 30 mmwith 1 zg/ml of LPS and then incubated for 60 mm with the mesotheialcells. The nonadherent cells were removed by three washes withmedium.


includes regulation of B cell immunoglobulin production andstimulation of hepatic acute-phase protein synthesis duringinflammation [63].

Dialysis solution and phagocytic cell functionIn 1981, Duwe, Vas, and Weatherhead first showed that

incubation of human peripheral blood leukocytes in peritonealdialysis solution inhibited the production of reactive oxygenspecies, phagocytosis, and bacterial killing [64]. Further inves-tigation by Alobaidi et al in the mid-1980s showed that phago-cytic function of peritoneal macrophages was significantlysuppressed by a pre-incubation period of only 10 minutes inCDS. Incubation in CDS for 30 minutes completely abolishedphagocytosis even though the cells remained viable [65]. In thisstudy, the macrophages were pre-incubated in the dialysissolution for various periods of time and then replaced in anormal cell culture medium for one hour before phagocyticfunction was assessed. This experimental design also providedevidence that the suppressor effect of short exposure to CDSwas not reversible even one hour alter the macrophages werereturned to a "normal" environment. If these findings can beextrapolated to CAPD, the initial suppressor effect of the CDSmight continue despite the gradual normalization of the dialy-sate (pH and osmolality) during its dwell time in the peritonealcavity. Other investigators have corroborated these findingsand have presented evidence that incubation in unused dialysisfluid inhibits the generation of oxygen-free radicals by activatedperipheral blood leukocytes as well as by peritoneal macro-phages [66—68]. By the late 1980s, these studies had demon-strated a suppressor effect of unused dialysis solution on thevarious functions of the phagocytic cells. However, the identityof the chemical components responsible for the cytotoxicitywas still not firmly defined. Topley and co-authors studied thisissue in depth [69]. They incubated human peritoneal polymor-phonuclear leukocytes and peritoneal macrophages in an un-used pentoneal dialysis solution that contained 35 mM oflactate and had a pH of 5.3 and an osmolality of 347 mOsm/kgH2O (made hypertonic by the addition of 1.5% glucose). In aparallel set of experiments, similar cells were incubated in aphosphate-buffered saline solution (PBS). After an incubationperiod of 30 minutes, the chemiluminescence response toopsonized zymosan was determined. Values for chemilumines-cence response also were obtained with cells pre-incubated inPBS and supplemented either with 35 mM lactic acid or withglucose to yield an osmolality of 400 mOsm/kg; the pH of thesolutions was adjusted either to 5.3 or 7.3. Pre-incubation in theunused dialysis solution or in the acidic PBS supplemented withlactic acid had a strong and almost identical suppressive effecton the chemiluminescence response. But the glucose-supple-mented PBS did not interfere with the subsequent chemilumi-nescence response to opsonized zymosan. The authors con-cluded that none of the components of CDS individually, suchas high glucose, high lactate, or low pH (5.3), had any inhibitoryeffects on the production of oxygen radicals by phagocyticcells. The chemiluminescence response was strongly sup-pressed, however, when lactate and a low pH were concomi-tantly present in the incubation solution.

The influence of hyperosmolality and the high glucose con-tent of the dialysis solution on phagocytic cell function werealso investigated by Van Bronswijk and colleagues [66]. They

examined the ability of polymorphonuclear leukocytes, mono-cytes, and peritoneal macrophages to mount a respiratory burstand phagocytosis of bacteria when exposed to media containingglucose concentrations ranging from 0.5% to 8.0%. They founda significant inhibition of chemiluminescence only when thecells were exposed to a glucose concentration higher than 4.0%,and a detrimental effect on phagocytosis at glucose levels higherthan 8.0%. These data differ from those of Liberek and cowork-ers, who found a suppressor effect on polymorphonuclear cellphagocytosis when glucose was added to the incubation me-dium in concentrations of 2.7% and higher [70]. The reason forthe discrepancy in the results is not clear.

The mechanism by which the hyperosmolar solutions impairphagocytic cell function deserves further investigation. Westudied volume-regulating behavior of peritoneal macrophagessuspended in CDS. We found a rapid decrease in volume asexpected from hyperosmotic stress, and this initial shrinkagephase was followed by yet another swelling towards normal size(unpublished observation). The data demonstrate that the peri-toneal macrophages can regulate their volume when exposed tohypertonic dialysis fluid. Further investigations are needed todefine the mechanisms by which dialysis solutions containinghigh glucose concentrations have a detrimental effect on thefunction of phagocytic cells.

Peritoneal dialysis fluid and cytokine releaseWe recently examined the effect of peritoneal dialysis solu-

tion on the production of IL-i and TNFa by peritoneal macro-phages [71]. The cells were pre-incubated for various periods oftime in unused CDS (1.5% glucose), then returned to the cellculture medium, and subsequently stimulated with LPS. After ashort pre-incubation period of 15 minutes in the dialysis solu-tion, the cells had a 50% decrease in the release of TNFa; alterone hour, the activity of TNFa was undetectable. We observedthis suppressor effect of CDS on TNFa activity in the superna-tant and cell lysates. We suggested that the inhibitory effect ofCDS on TNF occurs at the production level and not at thesecretion level. To define the mechanism by which CDS mark-edly attenuates LPS-induced TNFa production by the macro-phages, we studied the effect of the dialysis solution on induc-tion of TNFa message. We measured LPS-induced TNFamRNA production in human peritoneal macrophages by PCRand in a macrophage-like cell line, P388D1, by Northern blot-ting. We found that treatment of the cells with CDS for only 5to 15 minutes was sufficient to cause a marked decrease in thelevel of TNFa mRNA in both the macrophages and the cell line.In additional experiments, we investigated the physicochemicalconstituents of the dialysis solution that are responsible for theinhibitory effect of CDS on TNFa production. Cells werepre-incubated for 45 minutes in the test solutions and thenstimulated with LPS. Total RNA was extracted, size fraction-ated by 1.4% formaldehyde agarose, blotted, and hybridizedwith 32P-labeled TNFa and J3 actin probes. Cells were pre-incubated in test solutions as follows (Fig. 2): CDS (lane 1),lactate-free CDS (lane 2), glucose-free CDS (lane 3), lactate-and glucose-free CDS (lane 4), glucose-free CDS with added0.2% mannitol (lane 5), RPM! medium (lane 6). We noted thatpre-incubation of LPS-stimulated macrophages in a medium towhich 1.5 g/dl glucose was added, and with a pH adjusted to5.2, did not affect either the TNF activity in the supernatants or

a S °!fl

"VT


28S-

1 2 3 4 5 6 1 2 3 4 5 6

1 8S-

TNFa 3-actin

Fig. 2. Effect of CDS constituents on theexpression of TNFa mRNA in themacrophage-like cell line P388D1 by Northernblot technique.

the level of the TNFa mRNA signal. By contrast, pre-incuba-tion of the cells in a medium acidified to a pH of 5.2 in thepresence of 35 mM of lactate caused a marked attenuation ofthe TNFa mRNA signal. These results conform with those ofTopley et al, who also found that a low pH and a high lactateconcentration play an important role in mediating the suppres-sor effect of CDS in macrophage-phagocytosis [69].

The fundamental mechanism of the suppressor effect of CDSon TNF production was further investigated in our laboratorywhen we looked at some aspects of signal transduction inpentoneal macrophages. A DNA binding protein, NF-KB, me-diates, in a broad array of cells, signal transduction betweencytoplasm and the nucleus [72, 73]. In most cells, the pre-formed NF-KB transcription factor is complexed with a cyto-plasmic inhibitor named 1KB. Induction of NF-,cB by cellularactivation involves the disassembly of the NF-KB/IKB complex,perhaps by phosphorylation of 1KB, allowing the migration ofthe transcription factor to the nucleus. The interaction ofNF-KB with the TNFa promoter might be required for efficientLPS-mediated transcriptional activation of the TNFa gene [74,75]. The suppressor effect of dexamethasone on TNFa tran-scription is associated with decreased NF-KB activity. Thesedata led us to investigate the pattern of NF-KB activity innuclear extracts prepared from macrophages. We used a gelretardation assay (Fig. 3). Nuclear extracts were isolated from:A Non-stimulated and LPS-stimulated P388D1 macrophage-likecell line, with or without pre-incubation in CDS for 15 or 30mm; B Non-stimulated and TPA+Ionomycin-activated EL4cells that were positive controls for the normal migrationpattern of NF-KB; and C Competition experiment using 100-foldmolar excess of unlabeled NF-KB binding oligonucleotide. Gelretardation assays were performed with a 32P-radiolabeledNF-KB binding oligonucleotide. Protein-oligonucleotide com-plexes were identified by autoradiography after electrophoreticseparation in 5% TBE acrylamide gels. As Figure 3 shows,NF-KB was detectable in the nucleus of the LPS-stimulatedmacrophages. After a short, 15-minute treatment of the cellswith CDS, however, activation of NP-KB by LPS was markedlysuppressed. Other nuclear factors such as AP-1 were notaffected by exposure of the cells to CDS. Thus it seemsreasonable to assume that the inhibitory effect of CDS on TNFarelease by activated pentoneal macrophages can be explainedat least partially by an effect of this cytokine at a transcriptionallevel.

Jones et al recently studied the effect of CDS on the capacity

A B C

Fig. 3. Effect of CDS on NF-,.B DNA binding by gel retardationassays. The arrows indicate the retarded bands of (a) non-specificbinding activity, (b) the active NF-KB p65-p5° heterodimer, and (c) theinactive p5° homodimer.

of mononuclear and polymorphonuclear leukocytes to releaseTNFa [76]. Like us, they detected a strong suppressor effect ofthe dialysis fluid on TNFa release. Their experiments on thephysicochemical components responsible for the depressiveeffect of the dialysis solution revealed that the low pH andlactate exhibited a strongly depressive effect on TNFa. Unlikeour results, however, a mild but significant suppressor effectwas seen when they added only glucose to the incubationsolution at concentrations as low as 1.5%. This contradiction ofthe effect of glucose on TNFa release could be due to the

z-Il

w

LPS — + + + — — —

CDS — — 15 30 — — —

TPA+Ionomycin — — — — — + +

cold probe +

a-.

c+


A B C

- —F'- HBSSpH 7.0

CDSpH7.0

different assay systems used and to the fact that, in Jorres'experiments, cells were incubated in the glucose-containingsolution for 2 hours, a much longer period than we used. In thatreport and subsequent others, these investigators also noted asuppressor effect of CDS on the release of leukotriene B4, apowerful chemoattractant metabolite of arachidonic acid [76,77].

We then investigated the effect of CDS on the release of IL-iby LPS-stimulated macrophages and found that this cytokine isalso suppressed by the dialysis fluid. Unlike TNFa, however,after only one hour of exposure to CDS, IL-i release declinedsignificantly in the culture medium. Furthermore, 1.5% glucosewas required in the incubation medium in addition to lactate anda low pH to obtain an inhibitory effect on IL-i release compa-rable to that in the experiments carried out with CDS. We didnote with interest that IL-i and TNFa, which share so manysimilar biologic functions, displayed such different levels ofsensitivity to the inhibitory effects of CDS. The reason for thisdifference is not clear and warrants further study.

Collectively, these observations support the view that thecombination of low pH and high lactate in the peritonealdialysis fluid is detrimental to the function of the peritonealmacrophages. As far as the roles of hypertonicity and highglucose concentration are concerned, it would seem that inincubation media containing 1.5 g/dl glucose, as in standarddialysis solution, macrophage function is not compromised, buthigher concentrations of glucose might be detrimental. Theavailable evidence, however, does not provide us with enoughinformation regarding the specific cellular mechanism(s) in-volved in the suppressor effect of the lactic acid and low pH ofthe dialysis solution on the function of phagocytes.

Cellular pH and toxicity of dialysis solutionsMany disparate biologic functions depend on the mainte-

nance of a normal cytoplasmic pH [78]. In patients treated withCAPD, the fine regulation of cytoplasmic pH of the pentonealcells can be disrupted due to their repeated exposure to the verylow pH of the commercial dialysis fluid. A profound defect wasobserved in the ability of phagocytosis to mount a respiratoryburst and to produce cytokines when peritoneal macrophages

were exposed to CDS at its marketed pH of 5.2; these func-tional defects were corrected by the adjustment of the dialysisfluid pH to 7.4 [69, 79, 80].

We hypothesized that CDS suppresses macrophage functionby causing a drop in the cytoplasmic pH. To verify thisassumption, we measured the cytoplasmic pH of peritonealmacrophages fluorometrically while they were exposed to ei-ther CDS or to a physiologic saline solution adjusted to a pH of5.2 [81]. We found that exposure of the macrophages to CDSwas associated with a rapid and profound reduction in intracel-lular pH of the macrophages; the level achieved was 0.5 unitslower than the cytoplasmic acidification induced by the physi-ologic saline solution (Fig. 4A).

After instillation of the dialysis solution into the peritonealcavity, the dialysate rapidly becomes much less acidic; itreaches a pH of 6.5 after 15 minutes and a pH of 7.0 afterapproximately 30 minutes of dwell time [82]. This increase indialysate pH is due both to the diffusion of bicarbonate from theblood into the peritoneal cavity and also to the mixing of thefresh dialysis solution with residual fluid in the peritoneal cavityat the end of a dialysis cycle. To simulate the compositionalchanges in the first 30 minutes of dwell time, we conductedexperiments exposing the macrophages to CDS in which the pHwas adjusted to 7.0 by the addition of NaOH. Again we foundthat at any pH in the incubation medium, the cytoplasmic pHwas always much lower in the cells exposed to CDS than in thecells exposed to the saline solution (Fig. 4B).

In commercial peritoneal dialysis solutions, lactate is theorganic ion used as a base substitute, a proportion of which isundissociated in the acid pH of CDS. Consequently, we as-sumed that the lactate in its undissociated state (that is, lacticacid) traverses the plasma membrane into the macrophages andcauses intracellular acidification. To confirm this hypothesis,we exposed the macrophages either to a lactate-containingsaline solution or to CDS of similar pH. In both cases, thereduction in the cytoplasmic pH was identical (Fig. 4C). Thus itwould seem that the diffusion of lactic acid constitutes the mainacid generation pathway of the macrophages when they areexposed to CDS.

7.2

I

7.2

6.8

6.4

6.0

5.6

5.2

7.0

HBSSpH7.4

6.8

6.6

6.4

6.2

7.2

7.0

6.8

6.6

6.4

6.2

5 minutes

— HBSS pH 7.4

CDS pH 7.0

BSS pH 7.0+ lactate

5 minutes 5 minutesFig. 4. io BCECF-loadedperitoneal macrophages were incubated at 37°C, and the time course of cytosolic pH (pHi) was recorded. The baselinepHi was determined in Hepes buffered saline solution (HBSS) at pH 7.4. The traces of pHi after exposure to either CDS or HBSS at pH 5.2 isshown in A and at pH 7.0 is shown in B. The traces after the addition of 40 mM of lactate to HBSS is shown in C. Traces represent one of fivesimilar experiments.


During incubation of macrophages in CDS, the initial reduc-tion of cytoplasmic pH always was followed by a gradualrecovery of the pH toward baseline (Fig. 4B). Approximately40% of pH recovery could be inhibited by amiloride, aninhibitor of the Na/H exchanger. The remaining cytoplasmicpH recovery was insensitive to amioride and was independentof sodium, but it could be inhibited by treatment of the cellswith dicyclohexylcarbodiimide, an H-ATPase blocker. Thesedata support the idea that the peritoneal macrophages possesstwo efficient proton-extruding mechanisms involved in thecytoplasmic pH recovery of macrophages acidified with thedialysis solution: a sodium/proton exchanger and a protonpump.

In certain cells such as hepatocytes, lymphocytes, and me-sangial cells, the sodium-dependent C11HC03 exchanger isimportant in protecting against cytoplasmic acidification [83,84]. The sodium dependent C1/HC03 exchanger, whichexchanges extracellular sodium and bicarbonate for intracellu-lar chloride, can be blocked by the stilbene derivative, 4-4'-diisothyocyanostilbene-2-2' disulfonic acid (DIDS). Bicarbonateaccumulates in the dialysate during dwell time in the peritonealcavity, and if this exchanger is present in the peritoneal macro-phages, it possibly could protect against the acidifying effect ofCDS.

To evaluate this assumption, we determined the cytoplasmicpH of the macrophages exposed to CDS buffered to a pH of 7.0with C02-HC03. The decrease in the cytoplasmic pH wassimilar to that in cells exposed to a bicarbonate-free CDSbuffered to a pH of 7.0. Moreover, the reduction in cytoplasmicpH was not affected when we added DIDS to the incubationmedia. As far as we can extrapolate these in-vitro findings to thein-vivo conditions inside the peritoneal cavity, we can concludethat the sodium-dependent C1/HC03 exchanger is not impor-tant in protecting macrophages against the acidifying effect ofcommercially available dialysate solution.

The finding that CDS produces acidification of the macro-phages led us to question whether the cytoplasmic acidificationand the impaired immune function of these cells are causallyrelated. Although no unequivocal evidence exists regarding thisrelationship, some data suggest that a decrease in cytoplasmicpH indeed does impair the immune function of phagocytic cells.Let me illustrate such a phenomenon.

It is well known that the activity of cellular enzymes dependson an optimal pH. The NADPH oxidase is an enzyme thatcatalyzes the reaction of 02 with NADPH to form NADP andsuperoxide; the optimal pH range for this enzyme activation is6.8 to 7.9 [85]. We can logically assume, therefore, that whenthe macrophage-cytoplasmic pH falls well below 6.5, exposureto CDS will limit the activation of the NADPH oxidase andthereby impair the generation of superoxide. In an elegantpublication, Rotstein and coworkers described the effect of areduced cytoplasmic pH on superoxide production by phago-cytes [85]. These researchers tried to determine why certainanaerobic bacteria, particularly Bacteroides species, increasedtheir virulence by secreting large amounts of short-chain fattyacids such as succinic acid into the acidic milieu of theinflammatory sites. The augmented virulence of these bacteriais accounted for by their suppressor effect on leukocyte func-tion. Rotstein and colleagues measured the cytoplasmic pH ofneutrophils after exposing them to a medium simulating the

micro-environmental conditions created by the anaerobic bac-teria: a high concentration of succinic acid and a pH adjusted to6.5 or less. The cytoplasmic pH of these cells was profoundlyreduced and was associated with an impaired capacity of thecells to generate superoxide. To establish that the impairment inneutrophil oxidative metabolism was caused by the reduction inthe cytoplasmic pH, these workers clamped the cytoplasmic pHof these cells at various levels and observed that clampingcytoplasmic pH to values below 6.0 was associated with adefective generation of superoxide anions. On the basis of theseresults, the authors concluded that neutrophil function wascompromised by the entry of succinic acid into the cells andrelease of protons from succinic acid, thereby acidifying thecytoplasmic compartment.

Interestingly enough, this scenario resembles the conditionsprevailing in the peritoneal cavity of patients treated withCAPD after instillation of CDS, in which lactic acid accumu-lates in the peritoneal macrophages and results in a profoundreduction in macrophage cytoplasmic pH. It is tempting for usto speculate whether the inhibitory effect of CDS superoxideproduction of peritoneal macrophages is caused by its acidify-ing effect and a decreased activity of NADPH oxidase.

Vitamin D metabolism and peritoneal host defense

Thus far, I have concentrated on the role of cytokines andintracellular pH of macrophages in peritoneal host defense. Inthe last decade, we have learned a great deal about thenon-endocrine effects of vitamin D, and I would like to closethis Forum by spending some time discussing the possible roleof vitamin D in peritoneal host defense. Under normal physio-logic conditions, the kidney is a major source of 1 ,25(OH)2D3,the most biologically active hormonal form of vitamin D. In theproximal tubule cells, the enzymatic conversion of 25-OH-D3 to1 ,25(0H)D3 is catalyzed by the enzyme 25,1 a hydroxylase, amitochondrial cytochrome P450-linked mixed oxidase [86, 87].This reaction is tightly regulated by the serum concentration ofPTH, phosphorus, and 1 ,25(OH)2D3. Like other steroid hor-mones, the effects of l,25(OH)2D3 are mediated by a nuclearreceptor molecule to which the hormone binds with high affinity(Kd = 10—

10 M) and selectivity compared with other vitamin Dmetabolites. The receptor-hormone complex interacts with aDNA sequence enhancing the transcription of 1 ,25(OH)2D3-responsive genes [881.

As I mentioned earlier, evidence is increasing for a biologicrole of l,25(OH)2D3 beyond its regulatory role in mineral ionhomeostasis [89—91]. This hormone also might be locally pro-duced at sites of tissue inflammation and thus might play a keyrole in modulating the function of immune cells. Clinical evi-dence supporting the role of 1 ,25(OH)2D3 as a modulator of theimmune response is found in the frequent incidence of infectionin patients with nutritional rickets due to abnormal phagocyticfunction [92, 93]. Tissue macrophages and circulating mono-cytes constitutively express receptors for l,25(OH)2D3, butresting lymphocytes require activation by mitogenic lectins orspecific antigens. In physiologic concentrations, 1 ,25(OH)2D3inhibits lymphocyte proliferation and IL-2 production but stim-ulates monocyte/macrophage function [94]. Stimulatory effectsof 1 ,25(OH)2D3 have been documented in macrophage func-tions such as: phagocytosis, formation of hydrogen peroxide


and cytotoxicity, expression of Fc receptors, and expression ofIL-1$ and TNFa genes [90, 91, 95—97].

Prehn et al recently investigated the mechanism of thestimulatory effect of 1,25(OH)2D3 on the synthesis of TNF inLPS-stimulated macrophages [97]. The authors demonstratedthat pre-incubation of the myelomonocytic cell line U-937 with1 ,25(OH)D3 markedly potentiates both TNFa mRNA expres-sion and protein production. The authors presented persuasiveevidence that 1 ,25(OH)2D3-induced TNFa gene expressionoccurs not only by a direct transcriptional effect but also byincreased expression of CD14 mRNA, the high-affinity bindingsite for the LPS-LPS binding protein complex. The LPS bindingprotein has been isolated from the serum and cloned [98].Lipopolysaccharide complexed to the LPS binding protein ismuch more potent than native LPS, and its target in themonocytes, the CD14 antigen, is up-regulated by 1 ,25(OH)2D3[97].

Several studies have shown that 1 ,25(OH)2D3, like INF7, isan important stimulatory factor in the secretion of hydrogenperoxide by monocytes/macrophages. Cohen and colleaguesdemonstrated that pre-incubation for 3 days with 1 ,25(OH)2D3caused a threefold increase in the secretion of hydrogen perox-ide in phorbol-myristate-acetate-stimulated human monocytes[95]. This stimulatory effect was similar in magnitude to thatoccurring when the cells were treated with INF7. No synergis-tic effect was observed, however, when the cells were simulta-neously treated with both agents. The monocytes' diminishedcapacity to produce H202 when the culture period was pro-longed beyond 7 days could be offset by exposure of the cells to1,25(OH)2D3. Levy and Malech recently investigated the mech-anism by which 1 ,25(OH)2D3 potentiates production of super-oxide by phagocytic cells [99]. Under the influence ofl,25(OH)2D3, human monocytes augment the expression of a 47kD protein, an important cytosolic molecule involved in theactivation of NADP oxidase.

It is possible that 1 ,25(OH)2D3 plays a special immunoregu-latory role in the unique pentoneal micro-environment ofpatients undergoing CAPD. Peritoneal macrophages of non-infected patients undergoing CAPD can metabolize 25-OH-D3to 1,25(OH)2D3 in vitro [100], and the synthesis of this metab-olite is increased in patients with peritonitis. Synthesis of1 ,25(OH)2D3 depends on the presence of 25-la hydroxylase inthe macrophages, as evidenced by the fact that this reactioncould be blocked when ketaconazole was added to the incuba-tion medium. These in-vitro findings could be of clinical signif-icance for CAPD patients. A decade ago, we showed thatpatients undergoing CAPD lose 25-OH-D3 into the dialysate[101]. We also detected loss of 1,25(OH)2D3 in patients in whomoral replacement returned the plasma level of this metabolite tonormal [102]. This loss of vitamin D metabolites is mainly dueto the permeability of the peritoneal membrane to the serumvitamin-binding protein, a 52 kD a2 globulin, to which the 25-OH metabolite is tightly bound in the circulation. Conse-quently, the presence of 25-OH-D3 in the dialysate could be aconstant source of supply of substrate for the intrapentonealsynthesis of l,25(OH)2D3 by the macrophages. Thus, thismetabolite synthesized in the peritoneal cavity or lost into thedialysate could act as an autocrine or paracrine factor, enhanc-ing the immune function of the peritoneal macrophages. Werecently published work supporting this supposition [103].

We found that the production of superoxide by PMA-stimu-lated peritoneal macrophages treated for a 24-hour period withl,25(OH)2D3 was markedly enhanced in cells obtained fromCAPD patients with a low incidence of peritonitis. However, ina group of CAPD patients with a high incidence of peritonitis,the macrophages displayed a different pattern of behavior: thestimulatory effect of vitamin D was seen only after 3 days ofincubation. The resistance of the macrophages to 1 ,25(OH)2D3in this group of patients was associated with a high level ofprostaglandin secretion by these cells and could be overcomeby the addition of indomethacin to the culture medium. Ourfinding of elevated prostaglandin secretion by macrophagesfrom patients undergoing CAPD and having a high incidence ofperitonitis concurs with results obtained in a study by Lamperiand Carozzi [104]. We currently are studying the effect of1 ,25(OH)2D3 on the synthesis of cytokines by peritoneal mac-rophages.

A recent publication, in which the function of peripheralpolymorphonuclear leukocytes and monocytes in patients onchronic hemodialysis was studied before and after oral replace-ment of l,25(OH)2D3, further supports our work. Hubel andcolleagues found that replacement of 1 ,25(OH)2D3 normalizedmonocyte functions such as superoxide generation and bacte-ricidal capacity [105]. Polymorphonuclear cells functioned nor-mally in these patients, and 1 ,25(OH)2D3 treatment conferredno benefit.

One should keep these findings in mind when decidingwhether to administer 1 ,25(OH)2D3 replacement in patientsundergoing CAPD in whom the frequent incidence of low PTHand a low bone turnover might contraindicate the administra-tion of this metabolite [106]. It is our practice to administer1a25(OH)2D3 in doses varying from 0.5—1.0 j.gIday to ourCAPD patients, most of whom are being treated with dialysisfluid containing a low concentration of calcium (1.5 or 1.25mMol). We can thereby maintain a normal ionized calciumconcentration in the serum. All our patients receive calciumcarbonate as a binder of phosphate.

Questions and answers

DR. JOHN T. HARRINGTON (Chief of Medicine, Newton-Wellesley Hospital, Newton, Massachusetts): Thank you, Dr.Chaimovitz, for an excellent presentation. Given my interest inacid-base physiology, I'd like to start the question-and-answerperiod by asking you to expand on your comments regardingthe pHi of peritoneal cells. What is your methodology, itsvalidity, and its reproducibility?

DR. CHALMOVITZ: We measured the cytosolic pH of theperitoneal macrophages fluorometrically after loading the cellswith the pH-sensitive fluorescent dye BCECF. This is a stan-dard technique for determining cytosolic pH. The peritonealmacrophages were obtained from a dialysate of an overnightexchange in peritonitis-free CAPD patients. More than 90% ofthe cells were positive to the macrophage/monocyte markerCD14 monoclonal antibodies by flow cytometry. Calibration ofthese cytosolic pHs was performed after releasing the dye bylysing the cells with Triton and then recording the fluorescentsignal at known pHs. The data presented in the figures arerepresentative studies of several experiments revealing almostidentical changes in the cystolic pH.


Da. HARRJNOTON: Are there structural or functional differ-ences between the abnormal peritoneal cells of the patientsundergoing CAPD and the mesothelial cells from patients witheither cirrhotic or malignant ascites?

DR. CHAIM0vITz: I am aware of one study of patients notundergoing CAPD that characterized the function of peritonealmesothelial cells. This study, by Jonjic et al, demonstrated thatwhen activated, ascites-derived mesothelial cells from womenwith ovarian carcinoma synthesized cytokines and adhesionmolecules similarly to what has been shown with pentonealmesothelial cells of CAPD patients [58]. It would be interestingto further investigate the characteristics of peritoneal mesothe-hal cells in ascites of patients with salt-retaining disorders suchas cirrhosis of the liver.

DR. HARRINGTON: What is your prescription for a morephysiologic peritoneal dialysis solution?

DR. CHAIMovIrz: In-vitro as well as in-vivo studies havedemonstrated that the suppressor effect of currently availablecommercial dialysis fluid on the various functions of phagocyticcells could be corrected by raising the pH level of this fluid to amore physiologic value [69, 711. Thus, functions such as theability to generate a respiratory burst response, phagocytosis,killing of bacteria, and release of cytokines were corrected byraising the pH of the dialysis solution to 7.4.A low pH per se,however, does not seem to have a detrimental effect on thefunction of phagocytic cells. Lactate must be added to thesolution for the inhibitory effect of the incubation fluid toappear. Consequently, it is vital that more emphasis be put onresearch to produce a dialysis solution with a higher pH level.In my opinion, there is no decisive evidence proving that lactatein a solution of pH 7.4 is detrimental to the function ofperitoneal cells. I am not convinced, therefore, of the necessityof replacing this anion with bicarbonate in the dialysis solution.The osmotic agent currently in use in the dialysis solution mightbe toxic to mesothelial cells, and its absorption can produceobesity as well as disorders in lipid metabolism. Some prelim-inary accounts of a new osmotic agent, a hydrolysate from cornstarch, have reported promising results [107]. One of thedisadvantages in using this new compound however, is thepresence of very high levels of maltose in the serum.

DR. K. S. CHUGH (Postgraduate Medical Institute, Chandi-garh, India): Is there any difference between the propensity ofmesothelial cells of the peritoneum to develop fibrosis inpatients who have been on chronic hemodialysis compared withthose who are initially started on CAPD?

DR. CHAIMOVITZ: That is a very interesting question. I amnot aware of any publications that describe peritoneal biopsieswhen patients undergoing chronic hemodialysis are transferredto CAPD treatment. I understand that you also are concernedabout the effect of chronic uremia per se on the morphology ofthe peritoneal membrane. Dobbie described the morphology ofthe peritoneal membrane in patients who had end-stage renaldisease before starting CAPD [19]. These biopsies revealedminor changes such as cytoplasmic inclusions in the uremicperitoneal mesothelial cells as compared to normal controls.

DR. NICHOLAS TOPLEY (Institute of Nephrology, Universityof Wales, College of Medicine, Cardiff, Wales, United King-dom): I'm interested in the time course of the fibrotic events:Are the rapid, atypical events that occurred in this patientrelated to the 33 months the patient already has been on CAPD?

That is, does the treatment itself sensitize the peritoneal mem-brane to rapid sclerotic changes?

DR. CHAIMOVITZ: I agree absolutely that the rapid andirreversible deterioration of the peritoneal function in thispatient is a rare clinical course for a patient on CAPD afterperitonitis. We can speculate that prior to the peritonitis, thispatient was already suffering from chronic damage to theperitoneum, caused by exposure to the unphysiologic dialysissolution. But this damage, if present, was not clinically evidentprior to the peritonitis. I do believe that long-term preservationof membrane function should be a major goal of researchprojects on the biocompatibiity of the dialysis solution.

DR. TOPLEY: Does treatment with vitamin D3 make theperitoneal macrophages more mature? Can one measure sur-face expression of markers of maturity? I believe that we knowfrom many previous studies that peritoneal macrophages be-come increasingly immature with time on CAPD.

DR. CHAIM0vITz: That is a most interesting and challengingquestion. Exposure to 1 ,25(OH)2D3 reduces the proliferationand elicitates maturation of human tumor cell lines such asHL6O and U937 leukemic cell into cells with monocyte/macro-phage features [108, 109]. Thus, HL6O cells treated with1 ,25(OH)2D3 acquire morphologic features and cell-surfaceantigens of mature monocytes. Also, phagocytosis of S. aureusby HL6O cells was markedly stimulated by treatment with1 ,25(OH)2D3. Addition of 1 ,25(OH)2D3 at a concentration ofl0— M to culture of normal human marrow cells stimulatedmyeloid stem cells to differentiate into monocytes and macro-phages [109].

Let me return to your question. You are challenging us withthe possibility that the increasing immaturity of the pentonealmacrophages, as the duration in CAPD lengthens, can beexplained by an abnormal metabolism of vitamin D in thesepatients. Levels of 25(OH)D3 are low in CAPD because ofperitoneal loss of this metabolite. We now see more and morepatients in CAPD with very low plasma levels of I ,25(OH)2D3because nephrologists are reluctant to adequately replace thismetabolite, fearing the development of hypercalcemia and lowbone turnover disease. Hence, many patients undergoingCAPD are vitamin D deficient. The question remains whetherthe immaturity of the peritoneal macrophages is related tovitamin D deficiency; the answer is unclear and requires carefulexamination. We should remember, however, that the in-vitrodifferentiation-inducing effect of 1 ,25(OH)D2D3 requires supra-physiologic concentration of this vitamin D metabohite.

DR. CHARLES R. KLEEMAN (Cedars-Sinai Medical Center,Los Angeles, California): What is the effect of an osmoticallyactive substance other than glucose?

DR. CHAIMOvITz: Commercially available dialysis fluids uti-lize glucose as the osmotic agent. Studies in vitro and in vivosuggest that the high glucose content in the dialysis fluid inhibitsphagocytosis, bactericidal activity, chemiluminescence, andLTB4 synthesis of peripheral blood neutrophils and peritonealmacrophages [70, 76, 77]. The peritoneal absorption of glucosealso leads to metabolic disorders such as hyperlipidemia andobesity. Alternative osmotic agents now being studied areglycerol, amino-acid-based solutions, and glucose polymers.Lately, clinical studies using a glucose polymer with a molec-ular weight of 20,000 dalton have demonstrated significantultrafiltration in a prolonged dwell time of 12 hours in the


peritoneal cavity [107]. All these alternative osmotic agents,however, are still at the stage of clinical laboratory investiga-tion, and no unequivocal results have proved that they will beas effective as glucose.

DR. KLEEMAN: What happens to the phosphate content ofthe mesothelial cells during dialysis?

DR. CHAIMOVITZ: I am not aware of any studies on phos-phate concentration in mesothelial cells in patients on CAPD.

DR. KLEEMAN: What about content of TGFa and f3 andPDGF in the fluid if any? Also, what is the effect of substanceslike decorin, human proteoglycan, or antibodies to these cyto-kines?

DR. CHAIM0vITz: Shimokado et al demonstrated that acti-vated peritoneal macrophages secrete a mitogen for connectivetissue cells and that a large portion of this mitogenic activity isdue to a molecule similar to PDGF [30]. One can reasonablyassume that both PDGF and TGF/3 are significantly involved inthe pathogenesis of peritoneal fibrosis in CAPD. However,more research is required to support this assumption. Yung et alshowed that the dialysis effluent from CAPD patients containssubstantial amounts of proteoglycans and that the predominantglycosaminoglycan in the fluid is chondroitin/dermatan sulfate[110]. Omentum-derived mesothelial cells secrete the proteo-glycans decorin and biglycan [110, ill]. It is well known thatTGF/3 binds with high affinity to the core protein of these twoproteoglycans and that TGF bound to decorin will lose itsefficacy [112]. Since TGF up-regulates the synthesis ofdecorin, this binding process can represent an important mod-ulator of TGF/3 activity in the peritoneal cavity [113]. Again,more research is required to define the physiologic significanceof proteoglycan in the peritoneal fluid.

DR. KLEEMAN: Is there an experimental model of this prob-lem?

DR. CHAIM0vITz: Relatively few studies have been con-ducted using animal models to clarify the effect of the dialysissolution on the morphology, viability, and function of perito-neal cells. But Bos and colleagues did investigate whether thedialysis fluid could irritate the peritoneal cavity in rats [44].Using studies of the pattern of endogenous peroxidase activityof peritoneal macrophages, these researchers concluded that 24hours of exposure to the dialysis solution produced the appear-ance of an acute exudate in the peritoneal cavity of the animals.I agree that much effort should be made in this particular area.

DR. RAYMOND T. KREDIET (Academic Medical Center,Amsterdam, The Netherlands): The dialysate glucose concen-trations measured during the peritonitis suggest that the func-tion of the peritoneal membrane was not completely normalpreviously, because generally inflammation produces a net lossof ultrafiltration. This alteration necessitates the use of perito-neal dialysis fluids with a higher glucose concentration.

DR. CHALMOVITZ: The patient's chart shows that during thebout of peritonitis, he was still being treated with a dialysatefluid containing 1.5% glucose. Despite a marked decrease in theperitoneal ultrafiltration, his fluid balance was maintained byincreasing his urinary output via the admininstration of diuret-ics. I agree with your comment that during an episode ofperitonitis, a severe defect in ultrafiltration commonly requiresthe use of more hypertonic dialysis solutions. Experimentaldata show that a high concentration of glucose might have adetrimental effect on phagocyte function and a toxic effect on

mesothelial cells [22], so we must make every effort to avoidusing the 4.25% glucose dialysis fluid, especially when theperitoneum is already damaged by microorganic invasion.

DR. KREDIET: The PET results indicate an isolated loss of netultrafiltration, since the DIDO glucose ratio remained essen-tially within the normal range. This loss of ultrafiltration mightbe explained by the presence of pockets in the peritoneal cavityor by a loss of transcellular water transport. Rippe et al havedrawn attention to the possibility of the existence of ultra-smalltranscellular pores that allow water transport without solutetransport [114]. Loss of cellular small pores would abolish thedecrease of dialysate sodium concentration normally seen dur-ing the first hour of a PET test. My question is, did you measuredialysate sodium? But also allow me a further comment. A lowcreatinine clearance rate does not necessarily point to hypoper-meability of the membrane, because the clearance of low-molecular-weight solutes during CAPD dwells mainly dependon dialysate flow rate and thus will decrease as the drainedvolume decreases.

DR. CHAIMOVITZ: I agree that the marked decrease in theperitoneal ultrafiltration of this patient contributed to the clini-cal and biochemical evidence of underdialysis, since the PETtest did not reveal a defect in the peritoneal permeability tosolutes. You refer to the studies by Rippe et al, which sup-ported the concept of a "three-pore" model of membraneselectivity [114]. According to this model, the peritoneum canbe considered a membrane composed of a large number of smallpores (radius 40 to 60 A), a very small number of large pores(radius 200 to 300 A), and a transcellular ultra-small pore of 4—SA, which accounts for a substantial fluid flow driven by crys-talloid osmotic forces such as glucose during CAPD. Thetranscellular flow of fluid can explain the fall in the sodiumchloride concentration in the dialysate in the early phase ofdwell time. As I do not have available data on sodium concen-tration in the dialysis effluent in the first 2 hours of dwell time,I cannot confirm your interesting hypothesis that damage to theultra-small pores was involved in the peritoneal membranedysfunction in this particular patient.

DR. JAYSON RAPOPORT (Chief of Nephrology, Sheba MedicalCenter, Tel-Hashomer, Israel): You described how the perito-neal mesothelial cells can participate in the peritoneal inflam-matory response by producing inflammatory cytokines. If thesecells acted as antigen-presenting cells, the peritoneal inflamma-tory response might be greatly amplified. Do you have anyinformation on this point?

DR. CHAIMOVITZ: In our laboratory, Dr. Douvdevani col-lected data showing that human omentum-derived mesothelialcells expressed HLA-DR (unpublished data), and we are now inthe process of investigating the significance of this finding. Dr.Douvdevani, would you like to comment on this matter?

DR. A. DOUVDEVANI (Nephrology Laboratory, Soroka Med-ical Center): We found that human omentum-derived mesothe-lial cells strongly express HLA-DR after stimulation withy-interferon. The basal level of non-stimulated mesothelial cellswas very low. The expression of HLA-DR under inflammatoryconditions suggests that mesothelial cells serve as antigen-presenting cells, thereby contributing to the immune responseof the peritoneum.

DR. Z. KORZETS (Physician-in-Charge, Peritoneal Dialysis,Kfar Saba, Israel): Given the rapid equilibration of pH within


the peritoneal cavity, how does that equate with continuedcytotoxicity due to low pH?

DR. CHAIM0vITz: We studied the cytosolic changes in theperitoneal macrophages after acutely exposing them to perito-neal dialysis solution titrated to pHs of 6.0, 6.5, and 7.0 byadding NaOH. We tried to simulate the gradual alkalinization ofthe pentoneal fluid during the first 30 minutes of dwell time.Even after exposure to a dialysis solution titrated to a pH of 7.0,the cytosolic pH dropped and stabilized at values of 6.7. Thus,it would seem that in the first 30 minutes of dwell time, amarked acidification of the peritoneal cells occurred, therebyimpairing their function.

DR. TOMASZ LIBEREK (Institute of Nephrology, University ofWales School of Medicine, Cardff and Gdansk, Poland): Ithink I can answer Dr. Korzets' question concerning intracel-lular pH and pH equilibration during intraperitoneal dwell timea bit further. We have done similar studies in neutrophils andhave seen a similar drop in their intracellular pH duringincubation in Dianeal, a low-pH, high-lactate fluid. But whenwe transferred these cells to physiologic buffer, the intracellularpH rose again, apparently back to nonnal. However, thefunction of these neutrophils, at least as far as the respiratoryburst is concerned, was irreversibly damaged and did notrecover even as long as 4 hours later. I don't have a directanswer to your question, but our experiments suggest that therespiratory burst is affected at distal steps of intracellular signalpathway. We stimulated cells treated with Dianeal 5.2 withopsonized zymosan, PMA, FMLP, and calcium ionophoreA23 187. With none of these stimuli were we able to inducerespiratory burst activation as measured by chemiluminescenceresponse. With regard to the direct localization of this defect,Rotstejn et al have shown that NADPH oxidase, the centralenzyme in respiratory burst activation, is sensitive to changesin pH and that its activity is markedly inhibited at pH 5.0 [85].

DR. JOHN DONOHOE (Consultant Nephrologist, Beaumont!Mater Hospitals, Dublin, Ireland): Are we any nearer todeveloping "pH-friendly" peritoneal dialysate fluids, just as wehave moved away from acetate-based to bicarbonate-basedhemodialysis systems? Would lactate-free peritoneal dialysisfluid be advantageous?

DR. CHAIM0vITz: As I already mentioned, no in-vitro datasupport an inhibitory effect of lactate in phagocyte cell functionat a normal pH of 7.4 with concentrations such as those foundin the peritoneal dialysis solution (40 mEq!liter). As a first stepin improving the biocompatibiity of the dialysis solution, Isuggest using a dialysis solution containing a higher pH ratherthan substituting lactate for bicarbonate.

DR. CHARLES VAN YPERSELE (Professor of Nephrology,Université Catholique de Louvain, Brussels, Belgium): Theincidence of peritonitis has markedly decreased in recent yearsbecause of technical improvements. Could you give us somefigures on the current incidence (that is, per 100 patient years)of a loss of peritoneal surface as an impetus for treatmentmodality change?

DR. CHAIM0Vrrz: I agree that following improvements in theconnecting systems for CAPD there has been a marked declinein the incidence rate of peritonitis. However, much concernremains as to how to preserve peritoneal function in patients onlong-term CAPD. Approximately 15% to 19% of patients un-dergoing CAPD eventually have to change this mode of treat-

ment because of a loss of peritoneal function. Mactier recentlypresented data showing ultrafiltration failure rates in patients onCAPD: 3% of the patients on CAPD exhibited ultraffitrationfailure after one year, 10% after 3 years, and 31% after 6 years[115]. The majority of cases can be categorized in the type-Iultrafiltration failure category, in which rapid absorption ofglucose in the dialysis fluid limits the net transcapillary ultrafil-tration; only in a few patients was type-Il peritoneal dysfunc-tion observed, which was due to a decrease in the peritonealmembrane hydraulic permeability and/or peritoneal effectivesurface area.

Da. VAN YPERSELE: Rarely, some patients develop an in-creased ultrafiltration leading to their transfer to other treat-ment modalities. Could you comment on this type of complica-tion?

DR. CHAIMOVITZ: I have no personal experience with pa-tients undergoing CAPD who had an increased ultrafiltrationthat led to hypovolemia. I have followed patients on CAPD withend-stage renal failure due to medullary cystic disease of thekidneys and urinary salt losing. These patients were able tomaintain fluid balance by increasing their intake of water andelectrolytes.

DR. A. KAGAN (Consultant Nephrologist, Kaplan Hospital,Rehovot, Israel): What is the role of humoral factors such asIgG and lysozyme in the host defense mechanism in CAPDpatients? We have reported that the level of lysozyme in theperitoneal fluid of CAPD patients is high [116].

DR. CHAIMOVITZ: Humoral opsonizing factors such as IgGand C3 facilitate phagocytosis by coating microorganisms. Theconcentration of these opsonines in the dialysate was found tobe 40 to 50 times lower than in the serum [117]. However, therelationship between the low peritoneal opsonizing capacity inCAPD and the occurrence of peritonitis is not unanimouslyaccepted. Disagreement also exists among investigators regard-ing the significance of the low concentration of fibronectin in thedialysate and the frequency of peritonitis. I am familiar withyour paper that demonstrated lysozyme activity in the perito-neal effluent of CAPD patients [116]. However, the meaning ofthis finding deserves further investigation.

DR. JACOB GREEN (Cedars Sinai Hospital, Los Angeles,California): Was there any difference between Dianeal andnormal salt solution in the macrophages' ability to synthesize1 ,25(OH)2D3?

DR. CHAIMOVITZ: That is an important issue currently beinginvestigated by Dr. Shany in our laboratories. As you know,there are some controversial results regarding the suppressoreffect of acidosis on the renal conversion of 25(OH)D3 tol,25(OH)2D3 [118]. The likelihood exists that the cytosolicacidification of the peritoneal macrophages, which occurs afterexposure to the dialysis solution, limits the synthesis of1 ,25(OH)2D3 to these cells. We are in the process of studyingthis possibility.

Reprint requests to Dr. C. Chaimovitz, Head, Department of Neph-rology, Soroka Medical Center of Kupat Holim, P.O. Box 151, Beer-Sheva, 84101, Israel

References

1. TWARDOWSKI ZJ, NOLPH KD, KHANNA R, PROWANT BF, RYANLP, MOORE HL, NIELSEN MP: Peritoneal equilibration test. PentDial Bull 7:138—147, 1987


2. SLINGENEYER A, CANAUD B, MION C: Permanent loss of ultra-filtration capacity of the peritoneum in long-term peritoneal dial-ysis: an epidemiological study. Nephron 33:133—138, 1983

3. Ciozzi S, NAsINI MG, SCHELOTTO C, CAVIGLIA PM, SANTONI0, BARocci S, VERSACE F, CANTALUPPI A, SALIT M, LAMPERI S:Effects of Ca and 1 ,25(OH)2D3 on peritoneal immune-cellcytokine release and fibroblast proliferation in CAPD patients.Adv Pent Dial 5:111—120, 1989

4. LAMPERI S, CARozzI S, NASINI MG: Lympho-monokine disor-ders and peritoneal fibroblast proliferation in CAPD. Trans AmSoc Artf Intern Organs 32:35—38, 1986

5. Di.z-Buxo IA: Natural history of abnormal peritoneal permeabil-ity states, in Peritoneal Dialysis, edited by LA GRECA G, RoNcoC, FERIANI M, CHIARAMONTE S, CONZ P, Milan, Wichtig Editore,pp 177—180, 1988

6. WIDEROE TE, SMEBY LC, MJALAND 5, DAHL K, BERG KJ, AssTW: Long-term changes in transperitoneal water transport duringcontinuous ambulatory peritoneal dialysis. Nephron 38:238—247,1984

7. VERGER C: Relationship between peritoneal membrane structureand its permeability: clinical implications. Adv Pent Dial 12:87—95, 1985

8. SoMBoLos K, MCNAMEE P, MrrwALLI G: Autopsy findings inpatients treated by continuous ambulatory peritoneal dialysis.Pent Dial Bull 6:130—135, 1986

9. VERGER C, BRUNSCHVICG 0, LE CHARPENTIER Y, LAVERGNE A,VAUTELON I: Structural and ultrastructural peritoneal membranechanges and permeability alterations during CAPD. Proc Eur DialTransplant Assoc 18:199, 1981

10. KREDIET RT: Peritoneal transport characteristics of water, lowmolecular weight solutes and proteins during long term continuousambulatory peritoneal dialysis. Pent Dial Bull 6:61, 1986

11. SLATER DN, COPE HG, RAFTERY AT: Peritoneal ultrafiltrationafter chronic exposure to dialysis fluid. Pent Dial mt 12:48—50,1992

12. STRUIJK DG, KREDIET TR, KOOMEN GCM, HOEK FJ, Boascno-TEN EW, REIJDEN VAN DE HJ, Aiusz L: Functional characteris-tics of the peritoneal membrane in long-term continuous ambula-tory peritoneal dialysis. Nephron 59:213—220, 1991

13. OTA K, MESIMA M, WATANARE N, NAGAMURA S: Functionaldeterioration of the peritoneum, does it occur in the absence ofperitonitis? Nephnol Dial Transplant 2:30—33, 1987

14. DOBBIE JW, ZAKI MA: The ultrastructure of the parietal perito-neum in normal and uremic man and in patients on CAPD, inFrontiers in Peritoneal Dialysis, edited by MAHER JF, WINCHES-TER JF, New York, Field, Rich and Ass, 1986, pp 3—10

15. Di PAOLO N, SACCHI G, BUONCRISTIANI U, Rossi P, GAGGIOrrIE, ALESSANDRINI C, lanA L, Pucci AM: The morphology of theperitoneum in CAPD patients, in Frontiers in Penitoneal Dialysis,edited by MAHER IF, WINCHESTER JF, New York, Field, Rich andAss, 1986, pp 11—19

16. RONCO C: Pathophysiology of ultrafiltration in peritoneal dialysis,in Penitoneal Dialysis, edited by LA GRECA 0, RONCO C, FERIANIM, CHIARAMONTE 5, CONZ P, Milan, Wichtig Editore, 1991, pp71—84

17. DOBBIE JW: New concepts in molecular biology and ultrastruc-tural pathology of the peritoneum: their significance for peritonealdialysis. Am J Kidney Dis 15:97—109, 1990

18. Di PAOLO N, SACCHI G, DE MIA M, GAGGI0rrI E, CAPOTONDOL, Rossi P. BERNINI M, Pucci AM, IBBA L, SABATELI P,ALESSANDRINI C: Morphology of the peritoneal membrane duringcontinuous ambulatory dialysis. Nephron 44:204—211, 1986

19. DOBBIE JW, LLOYD JK, GALL CA: Categorization of ultrastruc-tural changes in peritoneal mesothelium, stroma and blood vesselsin uremia and CAPD patients. Adv Pent Dial 6:3—10, 1990

20. G0TL0IB L, SHOSTAK A: Peritoneal ultrastucture, in PeritonealDialysis, edited by NOLPH KD, Dordrecht/BostonlLondon,Kluwer Academic, 1989, pp 67—95

21. Bos HI, MEYER F, DE VELD J, BEELEN HJR: Peritoneal dialysisfluid induces changes of mononuclear phagocyte proportions.Kidney mt 36:20—26, 1989

22. LiBEREK T, TOPLEY N, MISTRY CD, COLES GA, MORGAN T,

QUIRK RA, WILLIAMS JD: Cell function and viability in glucosepolymer peritoneal dialysis fluid. Pent Dial mt 13:104—111, 1992

23. DOBBIE 1W: Pathogenesis of peritoneal fibrosing syndromes (scle-rosis peritonitis) in peritoneal dialysis. Pent Dial mt 12:14—27,1992

24. LAMPERI 5, Cozzi S, NASINI MG: Calcium antagonists im-prove ultrafiltration in patients on continuous ambulatory perito-neal dialysis (CAPD). TransAm Soc Artif Intern Organs 10:657—663, 1987

25. BltnBoRowlcz A, BALASKAS E, OREOPOULOS GD, VAS M, MAR-TIS L, SERKES KD, OanoPouLos DG: Interactions of cells fromperitoneal dialysate with mesothelial cells and fibroblast in cul-ture. Adv Pent Dial 8:7—10, 1992

26. WILLIAMS RS, RosSi AM, CHEGINI N, SCHULTZ G: Effect oftransforming growth factor on postoperative adhesion formationand intact peritoneum. J Surg Res 52:65—70, 1992

27. KOVACS EJ, FIBROGENIC CyT0KINEs: The role of immune medi-ators in the development of scar tissue. Immunol Today 12:1,17—23, 1991

28. LEROY EC, TROJANOWSKA MI, SMITH EA: Cytokines and humanfibrosis. Eur Cytokine Net 1:215—219, 1990

29. BORDER WA, RUOSLAHTI E: Transforming growth factor in dis-ease: the dark side of tissue repair. J Clin Invest 90:1—7, 1992

30. SHIM0KAD0 K, RAINES EW, MADTE5 DK, BAlutarr TB, BEN-DIrr liP, Ross R: A significant part of macrophage-derived growthfactor consists of at least two forms of PDGF. Cell 43:277—286,

198531. SELGAS R, L0PEZ-RIVAs A, MIRANDA B, MUNOZ J, Ririor C,

BoluuGoGo F, M0LINA 5, SANCHEZ-SICILIA L: Characterization

of the mitogenic-induced capacity of peritoneal effluent on humanand mice fibroblasts in culture. Nephnol Dial Transplant 6:44—50,1991

32. MAIORCA R, VONESH EF, CAVALLI PL, DE VECHI A, GIAN-GRANDE A, LA GRECA G, SCAiu'IoNI L, BRAGANTINI L, CANCAR-INI GC, CANTALUPPI A, CASTELNOVO C, CASTIGLI0NI A, P01-sErri PG, VIGLINO G: A multicenter selection-adjustedcomparison of patient and technique survivals on CAPD andhemodialysis. Peril Dial Int 11:118—127, 1991

33. GOKAL R, JAKUBOWSKI C, KINK 0, HUNT L, BOGLE S, BAILLODR, MARSH F, Ooo C, OLIVER D, WARD M, WILKINSON R:Outcome in patients on continuous peritoneal dialysis and hemo-dialysis: 4 years analysis of a prospective multicentre study.Lancet 2:1105—1 109, 1987

34. MAIORCA R, VONESH E, CANCARI GC, CANTALUPPI A, MANILIL, BRUNORI G, CAMERINI C, FELLER P, STRADA A: A six-yearcomparison of patient and technique survival in CAPD and HD.Kidney mt 34:518—524, 1988

35. MAIORCAR,CANCARIGC, BRUVNORI G, CAMERINI C, MANILIL:Morbidity and mortality of CAPD and hemodialysis. Kidney Int43(suppl 40):S4—S15, 1993

36. DE FIJTER CWH, VERBRUGH HA: Immunological response todifferent peritoneal dialysis solutions and schedules, in PenitonealDialysis, edited by LA GRECA G, RoNco C, FERIANI M, CHIAR-AMONTE 5, CONZ P, Milan, Wichtig Editore, Milan, 1991, pp179—184

37. PETERSON PK, GAZIANO E, SUH HJ, DEVALON M, PETERSON L,KEANE WF: Antimicrobial activities of dialysate-elicited andresident human peritoneal macrophages. Infect Immunol 49:212—218, 1985

38. HARVEY DM, SHEPPARDKJ, MORGAN AG, FLETCHER J: Effect ofdialysate fluids on phagocytosis and killing by normal neutrophils.J Clin Microbiol 8:1424—1427, 1987

39. GOLDSTEIN CS, BOMALASKI JS, ZURIER RB, NEILSON EG, Dou-GLAS SD: Analysis of peritoneal macrophages in continuousambulatory peritoneal dialysis patients. Kidney hit 26:733—740,1984

40. DAVIES SI, SUASSUNA J, OGG CS, CAMERON IS: Activation of

immunocompetent cells in the peritoneum of patients treated with

CAPD. Kidney Int 36:661—668, 198941. Bos HJ, SRUIJK DO, TUK CW, BEELEN RHJ: Characterization of

peritoneal cells from continuous ambulatory peritoneal dialysis(CAPD) patients: an immunological and functional study. J LeukBiol46:313, 1989


42. LEwIs S, HOLMES C: Host defense mechanisms in the peritonealcavity of continuous ambulatory peritoneal dialysis patients. PerilDial mt (first of Iwo parts) 11:14—21, 1991

43. Bos HJ, MEyER F, VELD JC, BEELEN RHJ: Peritoneal dialysisfluid induces changes of mononuclear phagocytic proportion.Kidney mt 36:20—26, 1989

44. Bos HJ, VAN BRoNswuK H, HELMERHOR5T TJM, OE PL, HOEF-5MIT ECM, BEELEN RHJ: Distinct subpopulation of elicitedhuman macrophages in peritoneal dialysis patients and womenundergoing laparoscopy: a study on peroxidatic activity. J LeukBiol 43:172—178, 1988

45. Bos HJ, SRUUK DG, TUK CW, DE VELD JC, HELERHOR5T TJM,HOEF5MIT ECM, ARI5z L, BEELEN RHJ: Peritoneal dialysisinduces a local sterile inflammatory state and the mesotheial cellsin the effluent are related to the bacterial peritonitis incidence.Nephron 59:508—509, 1991

46. McGIU300R Si, BROCHE JH, Btuous JD, JUNOR BJR: Bactericidalactivity of peritoneal macrophages from CAPD patients. NephrolDial Transplant 2:104—108, 1987

47. HOLMES Ci: CAPD-associated peritonitis: an immunological per-spective on causes and interventions. Contrib Nephrol 86:73—90,1990

48. STYLIANOU E, JENNER LA, DAVIES M, CoLts GA, WILLIAMS JD:Isolation, culture and characterization of human peritoneal me-sothelial cells. Kidney mt 37:1563—1570, 1990

49. DOUVDEVANI A, RAPOPORT J, KONFORTY A, SHMUEL A, OVNATA, CHAIMOvITZ C: Human peritoneal mesothelial cells (HPMC)synthesize interleukin-la (IL-I a) and /3: synergistic induction bylipopolyysaccharide (LPS) and tumor necrosis factor (TNFa) orLPS and IL-la. JAm Soc Nephrol 3:416, 1992

50. SCHINDLER R, LONNEMAN 0, SHALDON S, KOCH KM,DINARELLO CA: Transcription, not synthesis, of interieukin-l andtumor necrosis factor by complement. Kidney In: 37:85—93, 1990

51. DINARELLO CA: Interieukin-l and tumor necrosis factor and theirnaturally occurring antagonists during hemodialysis. Kidney ml42(suppl 38):568—577, 1992

52. KOLB H, KOLS-BACHOFEN V: Nitric oxide: a pathogenetic factorin autoimmunity. lmmunol Today 13:157—160, 1992

53. ZIMMERMAN GA, Paascorr SM, MCINTYRE TM: Endothelial cellinteractions with granulocytes: tethering and signaling molecules.Immunol Today 13:93—99, 1992

54. PARDI R, INvERARDI L, BENDER JR: Regulatory mechanisms inleukocyte adhesion: flexible receptors for sophisticated travelers.Immunol Today 13:224—230, 1992

55. MANTOvANI A, DEJANA E: Cytokines as communication signalsbetween leukocytes and endothelial cells. Immunol Today 10:370—375, 1989

56. SMITH CW, KIsrnMoTo TK, ABBA55 0,HUGHESB, ROTHLEINR,MCINTIRE LV, BUTCHER E, ANDERSON DC: Chemotactic factors

regulate lectin adhesion molecule 1 (LECAM-1)-dependent neu-trophil adhesion to cytokine-stimulated endothelial cells in vitro. Jdin Invest 87:609—618, 1991

57. ROTHLEIN R, WEaNER C: Role of intercellular adhesion mole-cule- I in the inflammatory response. Kidney mt 41:617—619, 1992

58. JONJIC N, PERI G, BERNASCONI 5, SCIACCA FL, COLOTTA F,PELICCI PG, LANFRANCONE L, MANTOVANI A: Expression ofadhesion molecules and chemotactic cytokines in cultured humanmesothelial cells. J Exp Med 176:1165—1174, 1992

59. ZEILLEMAKER AM, HOYNCK VON PAPENDRECHT AAGM, RoosD, LEGUIT P, VERERUGH HA: Neutrophil adherence and migra-tion through human mesothelial cell layer. Peril Dial mnt 13:30,1993

60. TOPLEY N, ZARIN B, JORRES A, WESTWICK J, DAVIES M, dOLESGA, WILLIAMS JD: Human peritoneal mesothelial cells (HPMC)synthesize interieukin-8 (IL-8): synergistic induction by interieu-kin-i and tumor necrosis factor (TNFa). Am J Pathol 142:1876—1886, 1993

61. BAGGIOLINI M, WALZ A, KUNKEL SL: Neutrophil-activatingpeptide-l/interieukin 8, a novel cytokine that activates neutro-phils. J Clin Invest 84: 1045—1049, 1989

62. TOPLEY N, Jotans A, LUTTMAN W, PETERSEN MM, LANG Mi,THIERAUCH KH, MULLER C, COLES GA, DAVIES M, WILLIAMS

JD: Human peritoneal mesothelial cells synthesize interieukin-6:induction by IL-i /3 and TNFa. Kidney Int 43:226—233, 1993

63. KI5HIM0T0 T: The biology of interleukin-6. Blood 4:1—10, 198964. DUwE AK, VA5 SI, WEATHERHEAD JW: Effects of the composi-

tion of peritoneal dialysis fluid on chemiluminescence, phagocy-tosis and bactericidal activity in vitro. Infect Immunol 33:130—135,1981

65. ALOBAIDI HM, COLES GA, DAVIES M, LLOYD D: Host defence incontinuous ambulatory peritoneal dialysis: the effect of the dialy-sate on phagocytic function. Nephrol Dial Transplant 1:16—21,1986

66. VAN BRONSwUK M, VERBRUGH HA, HEEZIUS HCJM, VAN DERMEULEN J, OE PL, VERHOEF i: Dialysis fluids and local hostresistance in patients on continuous ambulatory peritoneal dialy-sis. Eur J Clin Microbiol Infect Dis 7:368—373, 1988

67. Bos Hi, VLAANDEREN K, VAN DER MEULEN J, DE VELD JC, OELP, BEELEN RHJ: Peritoneal macrophages in short dwell timeeffluent show diminished phagocytosis. Peril Dial In: 8:199—202,1988

68. INci BL, GUPTA DK, NAwAB ZM, ZHOU F-Q, RAHMAN MA,DAUGIRDAS iT: Suppression of neutrophil superoxide productionby conventional peritoneal dialysis solution. Int J Aruf Organs11:351—354, 1988

69. TOPLEY N, ALOBAIDI HM, DAVIES M, COLaS GA, WILLIAMS JD,LLOYD D: The effect of dialysate on peritoneal phagocyte oxida-tive metabolism. Kidney Int 34:404—411, 1988

70. LIBEREK T, TOPLEY N, ioants A, COLE5 GA, GAHL GM,WILLIAMS JD: Peritoneal dialysis fluid inhibition of phagocytefunction: effects of osmolality and glucose concentration. J AmSoc Nephrol 3:1508—1515, 1993

71. DOUVDEVANI A, RAPOPORT i, KONFORTI A, ZLOTNIK M, CHAI-MOVITZ C: The effect of peritoneal dialysis fluid on cytokinerelease: factors involved and time course. Pent Dial Int 13:112—117, 1993

72. Hovos B, BALLARD DW, BOHNLEIN E, SIEKIEvITz M, GREENEWC: Kappa B specific DNA binding proteins: role in the regula-tion of human interleukin-l gene expression. Science (Wash, DC)244:457—460, 1989

73. SCHRECK R, BAEUERLE PA: NP-KB as inducible transcriptionalactivator of the granulocyte-macrophage colony stimulating factorgene. Mol Cell Biol 10:1281—1286, 1990

74. COLLART MA, BAEUERLE PA, VA55ILLI P: Regulation of tumornecrosis factor alpha transcription in macrophages: involvementof four KB-like motifs and of constitutive and inducible forms ofNP-KB. Mol Cell Biol 10:1498—1506, 1990

75. SHAKHOV AN, COLLART MA, VA55ILLI P, NEDosPAsov SA,JONGENEEL CV: KB-Type enhancers are involved in lipopolysac-

charide-mediated transcriptional activation of the tumor necrosis

factor gene in primary macrophages. J Exp Med 171:35—40, 1990

76. J0RRES A, JoRItEs D, GAHL GM, KES5EL M, MULLER C, Korr-GEN E, SERKE 5, SCHULZ E, MAHIOUT A: Leukotriene B4 andtumor necrosis factor release from leukocytes: effect of peritonealdialysis. Nephron 58:276—282, 1991

77. J0RRE5 A, JORRES D, TOPLEY N, GAHL GM, MAHIOUTA:Leukotriene release from peripheral and peritoneal leukocytes

following exposure to peritoneal dialysis solutions. Nephrol DialTransplant 6:495—501, 1991

78. GANAPATHY V, LEIBACH FH: Protons and regulation of biologicalfunctions. Kidney Int 40(suppl 33):54—Sl0, 1991

79. GORDON DL, RICE iL, AVERY VM: Surface phagocytosis andhost defence in the peritoneal cavity during continuous ambula-tory peritoneal dialysis. Eur J Clin Microbiol Infect Dis 9:191—197,1990

80. MANAHAN Fi, INC BL, CHAN JC, GUPTA DK, ZHOU FQ, PAL I,RAHMAN MA: Effects of bicarbonate-containing versus lactate-contalning peritoneal dialysis solutions on superoxide production

by human neutrophils, fin J Araf Org 13:495—497, 198981. CHAIMOVITZ C, RAPOPORT i, KONFORTI A, DOUVDEVANI A:

Commercial dialysis fluid causes intracellular acidosis in perito-neal macrophages. J Am Soc Nephrol 3:407, 1992

82. DIXON SR, MCKEAN WI, PRYOR JE, IRVINE ROH: Change inacid-base balance during peritoneal dialysis with fluid containinglactate ions. dim Sci 39: 51—60, 1990


83. BOYARSKY 0, GANZ MB, STERZEL RB, BORON WF: pH regula-tion in single glomerular mesangial cells. I. Acid extrusion inabsence and presence of HCO3. Am J Physiol 255(Cell Physiol24):C844—C856, 1988

84. ABDUL KARIM M, BATTLE D: Basic mechanisms of intracellularpH in lymphocytes. Semin Nephrol 11:3—15, 1991

85. ROTSTEIN OD, NASMITH PE, GRINSTEIN S: The bacteroids by-product succinic acid inhibits neutrophil respiratory burst byreducing intracellular pH. Infect Immunol 55:864—870, 1987

86. REICHEL H, KOEFFLER P, NORMAN AW: The role of the vitaminD endocrine system in health and disease. N Engl J Med 320:980—991, 1989

87. ADAMS JS: Hypercalcemia and hypercalciuria. Semin Resp Med13:402—410, 1992

88. OzoNo K, SONE T, PIKE JW: Perspectives: the genomic mecha-nism of action of I ,25-dihydroxyvitamin D3. J Bone Mm Res6:1021—1026, 1991

89. Tsouis CD, PROVVEDINI DM, MANOLAGAS SC: 1,25-dihydrox-yvitamin D3: a novel immunoregulatory hormone. Science 224:1438—1439, 1984

90. RIGBY WFC: The immunobiology of vitamin D. Semin Res Med13:402—410, 1992

91. HEwI5ON M: Vitamin D and the immune system. J Endocrino!132: 173—175, 1992

92. BAR-SHAVIT Z, N0FF D, EDELSTEIN S, MEYER M, SHIBOLET S,GOLDMAN R: 1-25 dihydroxyvitamin D3 and the regulation ofmacrophage function. Calcif Tissue Int 33:373—376, 1981

93. STOEDER J, KASAL P: Evaluation of phagocytosis in rickets. ActaPaediatr Scand 59:288—292, 1970

94. BHALLA AK, AMENTO EP, KRANE SM: Differential effect ofI ,25-dihydroxyvitamin D3 on human lymphocytes and monocyte/macrophages: inhibition of interleukin-2 and augmentation ofinterleukin-l production. Cell Immunol 98:311—322, 1986

95. COHEN MS, MESLER DE, SNIPES RG, GaRY TK: 1 ,25-dihydrox-yvitamin D3 activates secretion of hydrogen peroxide by humanmonocytes. J lmmunol 136:1049—1053, 1986

96. FAGAN DL, PREHN JL, ADAMS iS, JORDAN SC: The humanmyelomonocytic cell line U-937 as a model for studying alterationsin steroid-induced monokine gene expression: marked enhance-ment of lipopolysaccharide-stimulated interleukin-l messengerRNA levels by 1 ,25-Dihydroxyvitamin D3. Mol Endocrinol 5:179—186, 1991

97. PREHN JL, FAGAN DL, JORDAN SC, ADAMS JS: Potentiation oflipopolysaccharide-induced tumor necrosis factor expression by1 ,25-dihydroxyvitamin D3. Blood 80:2811—2816, 1992

98. SHUMANN RR, LEONG SR, FLAGGS OW, Gy PW, WRIGHT SD,MATI-IISON JC, TOBIAS PS, ULEVITZ RJ: Structure and function oflipopolysaccharide binding protein. Science 249:1429-1430, 1990

99. LEVY R, MALECH H: Effect of 1,25 dihydroxyvitamin D3 lipopoly-saccharide or lipoteichoid acid on the expression of NADPHoxydase component in cultured human monocytes. J Immunol147:3066—3071, 1991

100. SHANY S, RAPOPORT J, Zuiu I, GAVRIEL A, LAVI N, CHAIMOV-ITZ C: Metabolism of 25-OH vitamin D3 by peritoneal macro-phages from CAPD patients. Kidney mt 39:1005—1011, 1991

101. ALONI Y, SHANY S, CHAIMOVITZ C: Losses of 1,25-hydroxyvita-

mm D in pentoneal fluid: possible mechanism for bone disease inuremic patients treated with chronic ambulatory peritoneal dialy-sis. Miner Electrolyte Metab 9:82—86, 1983

102. SHANY 5, RAPOPORT J, Zuiu I, CHAIMOVITZ C: Losses of 1,25and 24,25 dihydroxyvitamin D3 in the peritoneal fluid of patientstreated with continuous peritoneal dialysis. Nephron 36:111—113,1984

103. LEVY R, KLEIN J, RUBINEK T, ALKAN M, SHANY S, CHAIMOVITZC: Diversity in peritoneal macrophage response of CAPD patientsto 1,25-dihydroxyvitamin D3. Kidney Int 37:1310—1315, 1990

104. LAMPERI 5, CAROZZI 5: Suppressor resident peritoneal macro-phages and peritonitis incidence in continuous pentoneal dialysis(CAPD). Nephron 44:219—225, 1986

105. HUBEL E, KIEFER T, WEBER J, METTANG T, KUHLMANN U: Invivo effect of 1,25 dihydroxyvitamin D3 on phagocyte function inhemodialysis patients. Kidney Int 40:927—933, 1991

106. SHERRARD DJ, HERCZ G, PEI Y, MALONEY NA, GREENWOOD C,MANUEL A, SAIPHOO C, FENTON SS, SEGRE GV: The spectrum ofbone disease in end-stage renal failure. Kidney mt 43:436—442,1993

107. MISTRY CD, MALLICK NP, GoIcL R: Ultraffitration with aniso-osmotic solution during long peritoneal dialysis exchanges.Lancet 2:178—182, 1987

108. DODD RC, COHEN MS, NEWMAN SL, Gity TK: Vitamin Dmetabolites change the phenotype of monoblastic U937 cells. ProcNat! Acad Sci USA 80:7538—7541, 1983

109. KOEFFLER HP, AMATRUDA D, IKEKAWA N, KOBAIASHI Y, DELUCA HF: Induction of macrophage differentiation of humannormal and leukemic myeloid stem cells by 1,25 dihydroxyvitaminD3 and its fluorinated analogues. Cancer Res 44:5624—5628, 1984

110. YUNG S, DAVIES M, WILLIAMS J, COLES 0: Human peritonealmesothelial cells (PMC) synthesize and secrete decorin and bigly-can. JAm Soc Nephrol 2:370, 1991

111. DI PAOLO N: Tissue culture and autologous implantation ofmesothelial cells, in Peritoneal Dialysis, edited by LA GRECA G,RONCO C, FERIANI M, CHIARAMONTE 5, CONZ P, Milano, Wich-tig Editore, 1991, pp 19—32

112. YAMAGUCHI Y, MANN D, RUOSLAHTI E: Negative regulation oftransforming growth factor-a by the proteoglycan decorin. Nature346:281—284, 1990

113. ROBERTS AB, MCCUNE BK, SPORN MB: TGF-: Regulation ofextracellular matrix. Kidney Int 41:557—559, 1992

114. RIPPE B, STELIN G, HARALDSSON B: Computer simulation ofperitoneal fluid transport. Kidney Int 40:315—325, 1991

115. MACTIER R: Ultrafiltration failure in CAPD (abstract). XIIth IntCong Nephro!, 1993, p 304

116. KAGAN A, BAR-KHAYIM Y, YARI A, FINARO M, RISNITZKY P:Serum and effluent lysozyme (muramidase) activity in CAPDpatients. Adv Pent Dial 5:154—156, 1989

117. HOLMES C, LEWIS 5: Host defence mechanisms in the peritonealcavity of continuous ambulatory peritoneal dialysis patients. PentDial Int (second of two parts) 11:112—117, 1991

118. CUNNINGHAM J, BIKLE DD, AVI0LI VL: Acute but not chronicmetabolic acidosis disturbs 25-hydroxyvitamin D3 metabolism.Kidney Int 25:47—52, 1984

peritoneal dialysis - core

Documents