Editorials

Dialyzer Reuse: What We Know and What We Don’t Know

Andre A. KaplanDivision of Nephrology, Department of Medicine and University of Connecticut Health Center,Farmington, Connecticut

ABSTRACT

Despite extensive clinical experience, the effects of different re-use procedures have not been fully evaluated. The available datasuggest that the effect of reuse on dialyzer performance dependsupon the type of chemicals employed, the membrane type, andthe size of the solute whose removal is being assessed. The effectof reuse on urea clearance is essentially defined by the residualcell volume with a total cell volume of > 80% associated with a

dialyzer clearance that is within 10% of its original value. Theeffect of reuse on large solute clearance can be dramatic, with theprocedure resulting in substantial changes in the b2-microglobulin clearance of different dialyzers. Of note is thelimited data available regarding the effect of reuse procedures ondialyzers processed more than 20 times.

Dialyzer reuse, first described by Shaldon et al. in1963 (1), is practiced in approximately 77% of all dialy-sis centers, caring for approximately 83% of all patientsin the United States (2). Of the germicides available,Renalin (peroxyacetic acid, acetic acid, and hydrogenperoxide) is used by 54% of units; formaldehyde, typi-cally used with bleach, by 38%; and glutaraldehyde by7%.

Despite this extensive clinical experience, the effectsof different reuse procedures on dialyzer performancehave not been fully evaluated. The available evidencesuggests that the following factors are relevant in anydiscussion of this topic: 1) the chemicals employed in thereuse procedure—Renalin, formaldehyde, bleach, andcitric acid all have unique effects on membrane perfor-mance; 2) the membrane type, its porosity, and its ten-dency to adsorb proteins; and 3) the size of the solutewhose removal is being assessed. There is clearly aninterest in small molecular weight clearance as measuredby urea kinetics and delivered Kt/V, but there is alsoincreasing interest in the removal of larger molecularweight substances such asb2-microglobulin (b2M). Thisarticle focuses on the effect of reuse on the removal ofsmall and large molecules.

Effect of Reuse on Urea Clearance

In keeping with current convention, the effect of reuseon the removal of small molecules will be focused onurea clearance. Early work by Farrell et al. (3) demon-strated that clearance of small molecules could be relatedto the residual amount of surface area in a reused dia-lyzer. Subsequently, Gotch (4) demonstrated that a re-

used dialyzer retaining 80% of its original cell volumewould maintain its urea clearance to within 5–11% of itsoriginal value. Given this relatively modest loss of ureaclearance, a total cell volume (TCV) of more than 80%was accepted as a standard for reused dialyzers (5), acriterion that was supported by subsequent data (6).

Since these studies were performed on low-flux dia-lyzers using the then standard blood flows of 200 ml/min, the question of whether an 80% TCV would ensureacceptable urea clearance under more modern operatingparameters was raised. Subsequent investigations gener-ally (but not uniformly) indicated that it does (7–9).Garred et al. (7) studied polysulfone dialyzers reused anaverage of 14 times with Renalin and operated at averageblood flows of 345 ml/min; urea and creatinine clear-ances were maintained and mass transfer coefficients re-mained relatively constant. Using high-flux dialyzers,blood flows of 400 ml/min, dialysate flows of 600 ml/min, and Renalin processing, Ouseph et al. (9) demon-strated that a residual cell volume of 80% did indeedmaintain Kt/V. Recent data from the HEMO study, in-volving a large experience with a wide variety of germi-cides and dialyzers, revealed only minor decreases inurea clearance, ranging from 1.1 to 2.9% per 10 reuses(10).

Not all data have supported the conclusion that an80% TCV standard assures acceptable urea clearances.Sherman et al. (11) reported data from an uncontrolledclinical practice setting involving 34 centers, 436 pa-tients, and a variety of reuse techniques. In centers usinga formaldehyde-based reuse procedure, Kt/V decreasedfrom 1.10 to 1.05 when the mean reuse number increasedfrom 3.8 to 13.8. Likewise, Murthy et al. (12) found thatincreasing the number of reuses resulted in substantialdeclines in urea clearance; after 20 reuses with formal-dehyde and bleach, urea clearance of polysulfone dialyz-ers decreased from 241 to 221 ml/min at a Qb of 300 andfrom 280 to 253 ml/min at a Qb of 400. More alarmingwas a report by Delmez et al. (13) demonstrating a 49%

Address correspondence to: Andre A. Kaplan, MD, FACP,MC-1405, Division of Nephrology, University of ConnecticutHealth Center, Farmington, CT 06030. E-mail: kaplan@nso.uchc.eduSeminars in Dialysis—Vol 13, No 5 (September–October)2000 pp. 271–274

271

decline in urea clearance after 15 reuses. This resultedfrom a dramatic misdistribution of dialysate in the dial-ysate compartment of a specific low-flux dialyzer. Thisanomaly went completely undetected by the TCV test,which is solely designed to evaluate the residual volumein the dialyzer’s blood compartment.

Thus the available data are not uniform, although areasonable conclusion is that urea clearance of reuseddialyzers is essentially defined by the residual cell vol-ume, with no systematic difference between reprocessingtechniques or with the type of membrane. One can alsoconclude that a residual cell volume of more than 80%ensures that urea clearance will remain reasonably closeto that of a new, unused dialyzer. The substantial declinein urea clearance despite a TCV of more than 80% [re-ported by Delmez et al. (13)]. is probably rare and relatedto a particular defect in a particular lot of a particulardialyzer.

Nonetheless, there are still issues to be addressed re-garding urea clearance and reuse. One must consider thatwhile the mean average number of reuses in the UnitedStates is 15, the mean maximum number of reuses is 36,with individual dialyzers being reused up to 192 times(2). Since few studies have evaluated dialyzers beyond20 reuses, many patients are currently being treated withdialyzers that are reused to a degree beyond the range forwhich there are available performance data. There maytherefore be a substantial number of dialyzers with TCVvalues closer to the 80% limit and with declines in ureaclearance closer to the “acceptable” 10%. A 10% declinein clearance (e.g., aKt/V of 1.17 instead of 1.30) isclinically significant, requiring, for example, that a treat-ment prescribed for 200 minutes be extended to 220minutes to maintain the baseline level of urea clearance.Given that the current TCV criteria allows for a decreasein urea clearance of this magnitude, reuse must be in-cluded in the multitude of other problems known to de-crease deliveredKt/V. With these considerations in mind,a modest proposal would be to raise the criteria for areused dialyzer so that the TCV is maintained at a valueof $90% in order to limit the acceptable loss of ureaclearance to a more modest 5% or less.

Effect of Reuse on Large MolecularWeight Clearance

In 1986 Rockel et al. (14), using a hemofiltrationmode, found that the sieving coefficient ofb2M in newpolysulfone dialyzers declined from 0.76 to 0.6 within 20minutes, decreasing further to 0.52 after 180 minutes.This decrease in sieving coefficient very likely reflectedprotein adsorption on the inner surface of the hollowfiber and the formation of a secondary membrane. Afterreuse three times with Renalin the dialyzers had only apartial regeneration of porosity with an increase from the180-minute “first-use” sieving coefficient of 0.52 to apost-Renalin initial coefficient of 0.58. These data sug-gest that reuse with Renalin allows at least some of thesecondary protein membrane to be retained, resulting ina modest decrease in the ability of the reused dialyzer toremoveb2M. This retained protein layering also blocks

complement activation sites and explains the increased“biocompatibility” of reused dialyzers processed withRenalin (15,16).

Thus with polysulfone dialyzers, the sieving coeffi-cient of larger molecular weight substances declines as aresult of secondary protein membrane formation on theinside of the hollow fibers. While changes in the totalcell volume are likely to influence middle and large mo-lecular weight solute clearance, their clearance is alsodetermined by the development of protein layering onthe inside of the membrane and whether the reuse pro-cedure can remove this layering.

With this as a background, the effect of bleach onreused polysulfone dialyzers can be easily contrastedwith that of Renalin. Early studies suggested that threereuses with bleach did not decreaseb2M removal bypolysulfone dialyzers nor affect the membrane’s substan-tial in vitro b2M adsorption (17). Diaz et al. (18) foundthat b2M clearance of polysulfone dialyzers increasedafter multiple reprocessing (#24) with bleach, with aprogressive decline in patients’ predialysis serumb2Mlevels. The notable but statistically insignificant increasein Kuf from 43.9 to 53.8 ml/min/mm Hg after 24 reuses,led the authors to conclude that the progressive decline inserum b2M levels resulted from a bleach-induced in-crease in the adsorptive capacity of polysulfone.

Subsequent data, however, suggest that at least part ofthe increasedb2M clearance with bleach reprocessing isdue to increased porosity and an enhanced sieving coef-ficient for b2M. In a study in which the entire dialysatewas collected and measured, we demonstrated a steadyincrease in dialysateb2M with increasing numbers ofbleach reprocessing (19). Unfortunately this increasedporosity involved other proteins as well, including albu-min, to the extent that protein losses increased to 20g/treatment when dialyzers were processed from 23 to 25times. Subsequent work by Krautzig et al. (20) revealedthat this increased porosity was the result of a leaching ofpolyvinylpyrrolidone (PVP) from the polysulfone mem-brane.

Thus, although bleach reprocessing removes the sec-ondary protein membrane which can decreaseb2M clear-ance, it also increases the overall porosity of the mem-brane to the extent that protein losses can become sub-stantial. Six months after removing bleach from the reuseprocedure we documented a substantial increase in ourpatients’ serum albumin levels (from 3.55 g/dl to 3.79g/dl) (19).

Fresenius, a major manufacturer of polysulfone dia-lyzers, now distinguishes their F80 brand of dialyzers;the F80A, more permeable to large molecules and reus-able with any reuse processing other than bleach, and theF80B, less permeable and designed to allow for repeatedbleach reuse without substantial protein loss. Gotch et al.(21) demonstrated that F80B polysulfone dialyzers re-used 20 times with bleach resulted in a high averageKoA for cytochrome C (surrogate forb2M) with onlyminimal protein loss. Similarly, Murthy et al. (12) re-ported that in F80B dialyzers reused with bleach,b2Mclearance increased from less than 5 ml/min at first use to23.6 ml/min after 20 reuses, again with only minimalprotein loss.

272 Kaplan

Thus bleach processing of polysulfone dialyzersmanufactured with PVP results not only in removal ofthe proteinaceous secondary membrane, but in an overallincrease in porosity that can improve the clearance ofmiddle molecules such asb2M.

Recently published data from the HEMO study hasprovided us with the most ambitious evaluation of theperformance of reused dialyzers (10). This NIH-sponsored study is a prospective, randomized, multi-center trial involving more than 45 dialysis units which isexamining the effects of differing prescriptions for Kt/Vand the ability of different membranes to removeb2M.The study compares low-flux dialyzers (b2M clearanceless than 10 ml/min) with high-flux dialyzers (b2M clear-ance greater than 20 ml/min). Reuse procedures utilizedin the participating study centers include Renalin, bleachcombined with either formaldehyde, glutaraldehyde, orRenalin, and heated citric acid.b2M clearances werecalculated from pre- and postdialysis serum values, amethod which has been demonstrated to yield artificiallyinflated values (see below). With the F80A dialyzer andRenalin processing,b2M clearance with an unused dia-lyzer was approximately 38 ml/min and remained rela-tively constant for up to 10 reuses. In contrast, when theCT190 dialyzer was processed with Renalin, initialb2Mclearances of 35–40 ml/min decreased steadily duringthe first nine reuses, and leveled off at approximately 15ml/min.

The opposite effect was found when bleach was partof the reprocessing procedure. With initial use, the F80Bhas ab2M clearance of 10–20 ml/min. When bleach isused with formaldehyde, porosity increases steadily,such that after 15–20 reuses,b2M clearance is more than50 ml/min. A similar increase inb2M clearance occurswhen bleach is used with Renalin, but not when bleach isused with glutaraldehyde, which only modestly increasesb2M clearance. The latter observation prompted the au-thors to speculate that reuse with glutaraldehyde resultsin a cross-linking of residual proteins left after bleachreprocessing. These cross-linked proteins would then ag-gregate into the microstructure of the membrane, de-creasing its effective pore size.

Repeated reuse of F80A dialyzers with heat and citricacid also yielded a modest increase inb2M clearance.These results confirmed our preliminary findings withheated citric acid, demonstrating a doubling ofb2Mclearance after 10 reuses without any substantial increasein net protein losses (22).

While the results of the HEMO study provide us withsubstantial information on the effect of reuse onb2Mclearance, several questions remain unanswered. Whathappens to total protein losses after 20 reuses (thestudy’s upper limit of reuse)? Since porosity increasessubstantially in an F80B dialyzer during the first 20bleach reuses, do net protein losses become excessiveafter 20 reuses? Given a mean maximum reuse numberof 36 in the United States (2), this issue deserves someattention.

Another compelling question is whether high- andlow-flux dialyzers actually differ substantially inb2Mremoval. Leypoldt et al. (23) comparedb2M clearance inhigh-flux dialyzers using pre- and postdialysis plasma

concentrations (as used in the HEMO study), with clear-ance determined using arteriovenous differences acrossthe dialyzer. Because of substantial postdialysis rebound(45% after 30 minutes), calculatedb2M clearances of 28ml/min were measured to be only 16 ml/min. Plasmalevels of b2M obtained 1 hour after treatment with a“high-flux” dialyzer (after a leveling of the posttreatment“rebound”) were not statistically different from theplasmab2M levels found after dialysis with a “low-flux”dialyzer. Thus the aforementioned changes inb2M clear-ance found in reused “high-flux” dialyzers may be amoot point.

Summary and Future Considerations

In summary, the effect of reuse on urea clearance isessentially defined by the residual cell volume. Giventhat a TCV of 80% allows for a 10% decline in ureaclearance, should this value remain an acceptable guide-line for dialyzer reuse? The effect of reuse on large soluteclearance can be dramatic with the procedure resulting insubstantial changes in theb2M clearance of differentdialyzers. Given that these effects depend on the numberof reuses, should dialyzers be reused for more than 20times without further information on the effect of theserepeated reuses on dialyzer performance?

References

1. Shaldon S, Silva H, Rosen SM: Technique of refrigerated coil preservationhemodialysis with femoral venous catheterization.Br Med J i:1716–1717,1963

2. Tokars JI, Miller ER, Alter MJ, Arduino MJ: National surveillance of dialy-sis associated diseases in the United States, 1995.ASAIO J44:98–107, 1998

3. Farrell PC, Eschbach JW, Vizzo JE, Babb AL: Hemodialyzer reuse: estima-tion of area loss from clearance data.Kidney Int5:446–450, 1974

4. Gotch FA: Mass transfer in reused dialyzers.Proc Dial Transpl Forum10:82–84, 1980

5. National Kidney Foundation: Revised standards for reuse of hemodialyzers.Am J Kidney Dis3:466–468, 1984

6. Gagnon RF, Kaye M: Dialyzer performance over prolonged reuse.ClinNephrol24:21–27, 1985

7. Garred LJ, Canaud B, Flavier JL, Poux C, Polito-Bouloux C, Mion C: Effectof reuse on dialyzer efficacy.Artif Organs14:80–84, 1990

8. Fleming SJ, Foreman K, Shanley K, Mihrshahi R, Siskind V: Dialyzer re-processing with Renalin.Am J Nephrol11:27–31, 1991

9. Ouseph R, Smith BP, Ward RA: Maintaining blood compartment, in dialyz-ers reprocessed with peracetic acid maintains Kt/V but not beta 2 micro-globulin removal.Am J Kidney Dis30:501–506, 1997

10. Cheung AK, Agodoa LY, Daugirdas JT, et al: Effects of hemodialyzer reuseon clearances of urea and beta-2 microglobulin.J Am Soc Nephrol10:117–127, 1999

11. Sherman RA, Cody RO, Rogers ME, Solanchick JC: The effect of dialyzerreuse on dialysis delivery.Am J Kidney Dis24:924–926, 1994

12. Murthy BVR, Sundaram S, Jaber BL, Perrella C, Meyer KB, Pereira BJG:Effect of formaldehyde/bleach reprocessing on in-vivo performances of highefficiency cellulose and high flux polysulfone dialyzers.J Am Soc Nephrol9:464–472, 1998

13. Delmez JA, Weerts CA, Hasamear PD, Windus DW: Severe dialyzer dys-function undetectable by standard reprocessing validation tests.Kidney Int36:478–484, 1989

14. Rockel A, Hertel J, Fiegel P, et al: Permeability and secondary membraneformation of a high flux polysulfone hemofilter.Kidney Int 30:429–432,1986

15. Dumler F, Zasuwa G, Levin NW: Effect of dialyzer reprocessing methods oncomplement activation and hemodialyzer-related symptoms.Artif Organs11:128–131, 1987

16. Westhuyzen J, Foreman K, Battistutta D, Saltissi D, Fleming SJ: Effect ofdialyzer reprocessing with Renalin on serum beta2 microglobulin and

273DIALYZER REUSE

complement activation in hemodialysis patients.Am J Nephrol12:29–36,1992

17. DiRaimondo CR, Pollak VE: Beta-2 microglobulin kinetics in maintenancehemodialysis: a comparison of conventional and high flux dialyzers and theeffects of dialyzer reuse.Am J Kidney Dis13:390–395, 1989

18. Diaz RJ, Washburn S, Cauble L, Siskind MS, Van Wyck D: The effect ofdialyzer reprocessing on performance and beta-2 microglobulin removal us-ing polysulfone membranes.Am J Kidney Dis21:405–410, 1993

19. Kaplan AA, Halley SE, Lapkin RA, Graeber CW: Dialysate protein losseswith bleach processed polysulphone dialyzers.Kidney Int47:573–578, 1995

20. Krautzig S, Mahiout A, Koch KM, Lemke HD: Reuse with oxidizing agents

leads to a loss of polyvinylpyrrolidone from polysulfone membranes [ab-stract].Blood Purif 12:176, 1994

21. Gotch F, Gentile D, Kaufman A, Levin N: Effects of reuse with peraceticacid, heat and bleach on polysulfone dialyzers [abstract].J Am Soc Nephrol5:415, 1994

22. Graeber GW, Halley SE, Lapkin RA, Graeber MW, Kaplan AA: Changes infilter permeability after reuse with citric acid and heat.Blood Purif15(suppl2):23, 1997

23. Leypoldt JK, Cheung AK, Deeter RB: Rebound kinetics ofb2-microglobulinafter hemodialysis.Kidney Int56:1571–1577, 1999

At the Present State of Our Knowledge

A most unusual retrograde pyelogram appeared in the Journal of Clinical Investigation in1955; the renal pelvis and calyces were clearly seen nestled in the shelter of the proximalfemur. In the accompanying article, Hume, Merrill, Miller, and Thorn (1) reported theirexperience at the Peter Bent Brigham Hospital in Boston with nine experimental renalhomotransplantations between 1951 and 1953. All kidneys after the first case were im-planted in the thigh, with the renal artery anastomosed to the profunda femoris artery andthe renal vein to the common femoral vein. The ureter was brought out percutaneously toa ureterostomy bag. This offered the advantage of sparing the recipient a nephrectomy,reducing operative trauma, allowing separate measurement of transplanted kidney ex-cretion, facilitating biopsy, and permitting early recognition of hematoma or infection.

Six of the donated kidneys were from patients who died during cardiovascular surgery.Three were from living, unrelated donors who required incidental nephrectomoies (distalureteral cancer and ventriculoureteral shunts for hydrocephalus). Donors and recipientsshared the same blood type in seven of the nine cases. The first transplant was performedat a referring hospital using a donor kidney from the ureteral carcinoma patient. Thekidney never functioned; the patient was transferred to the Brigham, where he died on the37th post-transplant day.

All other patients received their transplants at the Brigham. Four kidneys never functionedwell, and four functioned briefly; the patients died 19 to 101 days post-transplant. Mostreceived ACTH and cortisone. The final patient, transplanted in 1953, was a 26-year-oldphysician with chronic glomerulonephritis. The kidney was encased in a polyethylene bag,with vessels and ureter brought out through the bottom of the bag; no ACTH or cortisonewas used. Diuresis began on the 19th day and the BUN dropped slowly from 244 to 34mg%. The kidney ultimately failed, and the young physician died 176 days post-transplantation. The authors concluded: 1. the surgical procedure was technically fea-sible; 2. rejection seemed to be blunted by chronic renal failure; 3. ACTH and cortisone didnot appear to be beneficial in this series of patients.

At the present state of our knowledge, renal homotransplants do not appear to be justified inthe treatment of human disease. It is our conviction that this experience serves to indicatecertain important species differences in renal homograft response, and that as cellular andanimal experimentation progresses, constant cross-check with clinical experience is vitally es-sential to progress in this field of work.

These bold—and some would maintain, premature—transplantation experiments wereabandoned. The following year, Murray, Merrill, and Harrison successfully performed anidentical twin transplantation at the Brigham.

1. Hume DM, Merrill JP, Miller BF, Thorn GW: Experiences with renal homotransplantation in the human: reportof nine cases. J Clin Invest 34:327–382, 1955

274 Kaplan