The Wearable Artificial Kidney, Why and How:From Holy Grail to Reality

Victor Gura,* Claudio Ronco,† and Andrew Davenport‡

*Cedars Sinai Medical Center, UCLA, The David Geffen School of Medicine, Los Angeles, California,†Department of Nephrology Dialysis & Transplantation, San Bortolo Hospital, Vicenza, Italy, and ‡UCL Centerfor Nephrology, Royal Free & University College Medical School, London, United Kingdom

ABSTRACT

Once hemodialysis had become established as a treatment forchronic kidney disease, the early pioneers realized the limita-tions of the treatment, particularly in terms of the impactintermittent thrice weekly hemodialysis had on a patient’squality of life—not only time spent on dialysis and time trav-eling to and from treatment, but also dietary and fluid restric-tions. This led to the search for the holy grail—a wearablehemodialysis device (WAK), that would allow patients toreceive continuous treatment, while going on with the normalactivities of daily life. Such a device would not only provide

adequate solute clearances and control both electrolyteand acid–base status, but also improve blood pressure con-trol—all while allowing a liberal diet. Despite many attempts,to develop such a wearable artificial kidney, it is only recently,with the advent of microtechnologies, that it has been possi-ble to construct a truly wearable device, which can accuratelyregulate ultrafiltration and achieve adequate solute clear-ances. One such device has recently completed successfulhuman pilot studies, designed to test device reliability, safety,and efficacy.

Today routine outpatient dialysis is now regarded as awell-established technique, typically delivered in satellitedialysis units, private clinics, orminimal care centers,wellaway from the main hospital base. However, in the earlypioneering days, many obstacles had to be overcome,not only in the development of technology, but also theissues of reimbursement and patient eligibility. Besidesthe fundamental technological advances required todevelop dialyzers and dialysis machines (1), other keyadvances were required tomove treatment forward fromits restrictive beginnings limited to patients with acutekidney injury, to the provision of chronic dialysis treat-ments. These included reliable vascular access (2) andheparin anticoagulation toprevent circuit clotting.

Once the early pioneers had conquered these majorhurdles and could deliver successful hemodialysis topatients with chronic kidney disease (CKD), they real-ized the limitations of the therapy, particularly its effectson patients’ lifestyle. As such, the search for a dialysis

device that could be worn on a patient’s body can betraced back several decades to these early pioneers (3–7).The main rationale for these attempts to attain wear-ability, were patient convenience and improved qualityof life.

These initial attempts to develop a WAK were ham-pered because of the need for large amounts of freshwater for dialysate, the size and weight of then availableblood and dialysate pumps, and the lack of portableenergy sources to power these pumps. Sorbents wereused to reduce the amount of fresh dialysate required,but the typical sorbent cartridge weight of over 2 kgreduced mobility. None of these devices could be com-mercially developed. Others created wearable hemo-filters (3), typically using arterial or arterio-venousaccess (5), but to achieve adequate solute clearances thenencountered the formidable obstacle of replacing largeamounts of ultrafiltrate effluent with suitable replace-ment solutions for intravenous infusion or oral replace-ment. While these techniques proved useful in the reliefof fluid overload, they were not a commercially viablerenal replacement therapy (CRRT) for treating chronickidney failure patients.

Traditionally, hemodialysis patients have been offeredthrice weekly hemodialysis. Over time, with furtheradvances inmembrane technology, dialysis sessions con-tinued thrice weekly but times were shortened from8 hours down to a worldwide average of 4 hours (8).However, shortening the dialysis session, with a com-pensatory increase in ultrafiltration rate, leads toincreased frequency of intradialytic hypotension (9, 10),

Address Correspondence to: Victor Gura, MD, AttendingPhysician. Cedars Sinai Medical Center, Associate ClinicalProfessor of Medicine. UCLA, The Geffen School of Medi-cine, 9100 Wilshire Blvd. Suite 360W, Beverly Hills, CA 90212,or e-mail: vgura@cs.com.

Disclosures: Victor Gura is the Chief Medical Officer and adirector of Xcorporeal Inc.Seminars in Dialysis—Vol 22, No 1 (January–February) 2009pp. 13–17DOI: 10.1111/j.1525-139X.2008.00507.xª 2008 Copyright the Authors.Journal compilation ª 2008 Wiley Periodicals, Inc.

13

with corresponding cardiac stunning. There is nowmounting evidence that increased dialysis frequency andprolonging dialysis sessions improve patient quality oflife, and potentially increase life expectancy for CKDpatients (11–15). Increasing the duration of the dialysissession allows increased clearance of so called ‘‘middlemolecules,’’ and a slower, better tolerated rate of fluidremoval. The emerging data on the benefits of daily dial-ysis seem to indicate that frequent and incremental dialy-sis session time are key for improving the dismaloutcomes we currently achieve for CKDpatients.The concept that filtering blood for 9–12 hours ⁄week

can restore the health of CKD patients back to that of anormal person, in whom the native kidneys filter blood24 hours a day, 7 days a week, appears somewhat naive.Further complicating the time issue is the physical chem-ical fact that solutes of different size travel through dialy-sis membranes at different speeds, according to theirmolecular weight and charge. Thus, larger moleculesdiffuse slower, and may not adequately transit fromthe plasma water to the dialysate in the limited timeprovided.In addition, several uremic toxins [e.g., p-cresol,

reported to be associated with increased mortality (16)]are protein bound, and as such only a small amount isfree in plasma and available for removal. Time isrequired for protein bound toxins to re-equilibrate andraise the plasma level of free toxin, and permit additionalremoval. Continuous removal of the free fraction wouldcreate a gradient from bound to free fraction resulting inthe effective removal of toxins. Sorbents can increase theremoval of these protein-bound toxins, such as p-cresol(17).It has also been shown that phosphate, an indepen-

dent factor of all cause mortality in CKD patients, isonly efficiently removed when dialysis session time isextended (18). Therefore, it seems that more frequentand longer treatments may be far more physiologicalthan the current paradigm of 9–12 hours a week treat-ment. The advantages of daily extended dialysis are sum-marized in Table 1.

On the other hand, implementing more frequent andlonger, or daily dialysis programs is easier said thandone. There are major practical issues, including appro-priate manpower (nurses and technicians to supervisemore treatments in the dialysis units), building new dial-ysis facilities to cope with this expansion in dialysisrequirement, and the reluctance of governments andother payers to shoulder the expense of longer and morefrequent dialysis (19–22). In addition, patients may notwish to dialyze longer or more frequently and may beunable or unsuitable to dialyze at home.Even if governments and payers would agree to pay

for the additional costs of daily dialysis, executing such ataskwould take time. One alternative would be tominia-turize a dialysis machine and make it feasible for apatient to wear it.

Requirements for Continuous RenalReplacement Therapy

The various modes of continuous CRRT can deliversignificantly higher doses of dialysis, by treating patients24 hours a day, 7 days a week. However, CRRTmachines are not suitable to treat CKD patients, as theyrequire constant nursing time to treat a single patient. Inaddition, they are heavy, tethered to a wall electricaloutlet, and require large volumes of sterile replacementfluid or dialysate (23).Whilst treated, patients are unableto mobilize and perform their activities of daily life. Aminiaturized and wearable CRRT machine (Fig. 1A)could potentially solve these problems. In order to buildaWAK the following challenges had to be overcome:

Power Source

Traditional hemodialysis machines run onmains elec-tricity, with a back up heavy battery. So to be portable,a WAK must have a battery that, though small andlight, will provide enough energy to power all the neces-sary systems for a significant period of time to make theWAK independent of a fixed electrical outlet.

Dialysate

Standard hemodialysis therapy requires large volumesof fresh dialysate. The volume of fresh dialysate wouldrequire a huge weight burden that would render wear-ability impossible. Thus, a WAK requires a sorbent sys-tem which can purify and regenerate effluent dialysate,so avoiding the need for fresh dialysate. In addition,there is emerging evidence that dialysate should be ultra-pure and free not only of bacteria but also of toxins andpyrogens. To provide such quality dialysate, the WAKuses sterile 0.45% saline in the dialysate circuit and boththe tubing and sorbent systems are gamma sterilized.

Additives

The final dialysate is made by adding an electrolytesolution and bicarbonate to the dialysis water, usinga proportionating system in the dialysis machine.

TABLE 1. The potential benefits of extended daily dialysis leading

to improved outcomes in the treatment of CKD patients

Improved soluteclearances

No hyperkalemiaImproved appetite and nutritionLess bone disease and hyperparathyroidismEliminate the need for phosphate bindersImproved acid-base controlLess pruritusFewer sleep disturbancesReduced restless legsLess anemia and reduced ESA requirement

Improved volumecontrol

Appropriate sodium balanceImproved blood pressure controlDecreased use of anti hypertensive drugsReduced intradialytic hypotensionReduced risk of cardiac death and stroke

Improved serumalbumin

Fewer cerebrovascular eventsLower expected morbidity and mortality

Improved sleeppatterns

Less sleep apnea

14 Gura et al.

Similarly, theWAKwas designed to have two additionalpumps, one for a bicarbonate solution and a second foran electrolyte solution to be added to the dialysate com-partment (Fig. 1B).

Fluid Removal

Standard hemodialysis machines allow controlledultrafiltration. Thus, the WAK must have a volumetricpump to remove fluid at a physiological rate to avoidhemodynamic problems and yetmaintain euvolemia.

Ergonomic

For theWAK to be truly portable for prolonged peri-ods, it has to be light and ergonomically adapted to thebody contour so that it can be worn continuously with-out impinging on the patient’s ability to sleep, ambulateor perform the activities of daily life.

Pumps

Standard hemodialysis machines use large heavyroller pumps to propel blood and dialysate. A WAKbased on these pumps would be too heavy to be porta-ble. So a unique double channel pump was developed topropel both blood and dialsyate in opposite phase for

theWAK; it requires less energy than the standard dialy-sis machine roller pumps. The flow patterns generatedby this dual chamber pump differ from those of conven-tional blood pumps with an intermittent inversionof transmembrane pressures creating a ‘‘push pull’’mechanism that further improves convective transport.Push-pull mechanisms for increasing convective masstransport have been previously proposed, by using apiston pump to propel dialysate (24) but never commer-cially developed. The pump currently weighs 300 gramand is battery operated.

Safety

Commercial dialysis machines are equipped withsafety systems to prevent passage of air to the patient,and to stop the blood pump in case of a disconnection.Thus, theWAKwas developed to include a servomecha-nism with a bubble detector sensor placed after theblood pump, designed to stop blood flow if air bubbleswere detected in the blood circuit, and a second servo-mechanism to halt the ultrafiltration pump if the bloodflow stopped for any reason. In addition, the pulsatileblood pump had a self-limited capacity to generate nega-tive pressure for suction from the arterial side of the vas-cular access, such that significant negative pressurescould not be applied to the vascular access. Thus, any

(A)

(B)

Fig. 1. (A) Photograph of WAK. (B) Circuit diagram of WAK.

THE WEARABLE ARTIFICIAL KIDNEY 15

disconnection on the arterial side would result in cessa-tion of the blood pump.In addition to our development of a WAK,

others have proposed a peritoneal dialysis version (25),based on regeneration of the peritoneal dialysate effluent(26).

Laboratory and Animal Testing

When all these challenges were met, a battery oper-ated WAK was built in our laboratory (Fig. 1A). Theinitial prototypeWAK could be worn as a belt, weighedapproximately 10 lbs and was tested both in vitro and ina porcine uremic model where we demonstrated itssafety and efficacy (27). TheWAK, conceived as a wear-able CRRTmachine, was also configured as a hemofilterfor the treatment of fluid overload; when studied in fluidoverloaded animals, it achieved ultrafiltration rates ofup to 700 ml per hour, without difficulties or complica-tions (28). This application may be potentially useful inthe treatment of NYH class III and IV congestive heartfailure patients who develop a cardiorenal syndrome,given the current interest in the use of isolated ultrafiltra-tion in these patients (29).

Technical Characteristic of WAK

The WAK used in our preliminary studies, useda standard commercial 0.6 m2 high flux polysulfone dia-lyzer (Gambro Dialysatoren, Germany). The dialysatewas regenerated by using a series of three sorbent canis-ters, containing urease, activated charcoal, and bothhydroxyl zirconium oxide and zirconium phosphate.Patients were connected to the WAK by their usual vas-cular access for hemodialysis, either by central venouscatheters or native arteriovenous fistulae, and were anti-coagulated with unfractionated heparin, using theirstandard loading dose and maintenance dose as per atypical hemodialysis session, and the adjusted, aimingfor an aPTTr of 1.5–2.0. The mean blood flow around60 ml ⁄minute, with a dialysate flow of 50 ml ⁄minute.The ultrafiltrate rate was set according to clinical needand controlled by a volumetric pump. The total weightof the device was approximately 5 kg.

Human Studies

Following the successful animal trials, pilot studieswere performed in human subjects. In one study, iso-lated ultrafiltration was achieved safely and efficiently,without any side effects or complications, in six patientswith CKD (30). Subsequently, the WAK was tried in aproof of concept, feasibility study in eight humans forperiods varying between 4 and 8 hours (31). In thisstudy, all patients tolerated treatment with the WAKwithout symptoms or complaints. They were able tosleep, walk around, sit, eat, and drink without hin-drance. Urea clearance was 22.7 � 5.2 ml ⁄minute andcreatinine clearance was 20.7 � 4.8 ml ⁄minute, with an

hourly Kt ⁄v of 0.035. Although one catheter clotted andone patient had a needle dislodged, there were no com-plications attributable to the WAK; the safety devicesstopped blood and dialysate flows at the time of the inci-dents. The sorbent system generated CO2 bubbles in thedialysate circuit that partially interfered with the dialy-sate flow. This problem has now been successfullyresolved in our laboratory. Following the initial humanpilot studies, the initial WAK prototype is currentlybeing transformed into a finalized model that can becommercially developed.The WAK has been designed for continuous use to

deliver 168 h per week of dialysis, about 16 times morethan that currently usedwith standard thrice weekly out-patient hemodialysis. Effective and slow removal ofsodium and water will result in better control of hyper-tension and fluid overload (32), and may also eliminatethe traditional need for dietary restrictions of water andsalt (33). Similarly, the amounts of potassium and phos-phate removed should also result in the elimination ofdraconian dietary restrictions, andmaymake phosphatebinders obsolete (12, 14). These notions are supportedby the results obtained with daily (11–14) or prolonged(8, 11, 34, 35) dialysis.

Summary

Although it has taken more than 40 years to developa prototype of a truly wearable artificial kidney, the holygrail of the original pioneers in the field of dialysis, ourwearable device has successfully completed its pilotsafety and efficacy trials. The WAK could potentiallyprovide patients with a solution to increasing the dura-tion and frequency of dialysis therapies, without the cap-ital investments and nursing manpower needed today toprovide more frequent or prolonged dialysis to CKDpatients.Further clinical studies are now required to substanti-

ate the efficacy and safety of the WAK and obtain regu-latory approval. Provided this is accomplished, theWAK has the potential to become the standard of carefor dialysis in the future.

References

1. Kolff WJ: Artificial kidney, treatment of acute and chronic uremia.Cleve Clin Q 17:216–218, 1950

2. Quinton WE, Dillard DH, Cole JJ, Scribner BH: Possible improvementsin the technique of long-term cannulation of blood vessels. Trans AmSoc Artif Intern Organ 7:60–77, 1961

3. Bonomini V, Roggeri G, Inventors: hemodialysis and ⁄ or ultrafiltrationapparatus. US patent 4269708, 5 ⁄ 26 ⁄ 1981

4. Kolff WJ, Jacobsen S, Stephen RL, Rose D: Towards a wearable artifi-cial kidney. Kidney Int Suppl, S300–S304, 1976

5. Shaldon S, Beau MC, Deschodt G, Lysaght MJ, Ramperez P, Mion C:Continuous ambulatory hemofiltration. Trans Am Soc Artif InternOrgans 26:210–223, 1980

6. Murisasco A, Reynier JP, Ragon A, Boobes Y, Baz M, Durand C,Bertocchio P, Agenet C, el Mehdi M: Continuous arterio-venous hemo-filtration in a wearable device to treat end-stage renal disease. Trans AmSoc Artif Intern Organs 32:567–571, 1986

7. Senoo S, Otsubo O, Watanabe T, Yamauchi J, Yamada Y, Inou T,Takai N, Takahashi H, Fukui H, Kawata Y: The wearable artificial kid-ney: Development of a small blood pump. Jinkou Zouki 11:48, 1982

8. Chazot C, Jean G: Treatment time. Contrib Nephrol 161:154–161, 2008

16 Gura et al.

9. Davenport A, Cox C, Thuraisingham R: Achieving blood pressure tar-gets during dialysis improves control but increases intradialytic hypo-tension. Kidney Int 73:759–64, 2008

10. Davenport A: Intradialytic complications during hemodialysis. Hemo-dial Int 10:162–167, 2006

11. Kjellstrand CM, Ing T: Daily hemodialysis: History and revival of asuperior dialysis method. ASAIO J 44:117–122, 1988

12. Lindsay RM, Heidenheim AP, Leitch R, Ryan H, Kroeker A, Peters K,Workentin L, Nesrallah G, Prakash S, Kortas C; Daily ⁄ Nocturnal Dia-lysis Study Group: Short daily versus long nocturnal hemodialysis.ASAIO J 47:449–455, 2001

13. Buoncristiani U, Fagugli R, Quintaliani G, Kulurianu H: Rationale fordaily dialysis. Home Hemodial Int 1:12–18, 1997

14. Depner T: Why daily hemodialysis is better: Solute kinetics. Semin Dial12:462–471, 1999

15. Pierratos A: Effect of therapy time and frequency on effective soluteremoval. Semin Dial 14:284–288, 2001

16. Bammens B, Evenpoel P, Keulers H, Verbeke K, Vanrenterghem Y:Free serum concentrations of the protein-bound retention solutep-cresol predict mortality in hemodialysis patients. Kidney Int 69:1081–1087, 2006

17. Meyer TW, Peattie JW, Miller JD, Dinh DC, Recht NS, Walther JL,Hostetter TH: Increasing the clearance of protein-bound solutes byaddition of a sorbent to the dialysate. J Am Soc Nephrol 183:868–874,2007

18. Gallar P, Ortiz M, Ortega O, Rodrıguez I, Seijas V, Carreno A, Oliet A,Vigil A: Factors which influence phosphorus removal in hemodialysis.Nefrologia 27:46–52, 2007

19. Lockridge Jr RS: The direction of end-stage renal disease reimburse-ment in the United States. Semin Dial 17:125–130, 2004

20. Lockridge Jr RS, McKinney JK: Is HCFA’s reimbursement policy con-trolling quality of care for end-stage renal disease patients? ASAIO J47:466–468, 2001

21. Manns BJ, Johnson JA, Taub K, Mortis G, Ghali WA, Donaldson C:Dialysis adequacy and health related quality of life in hemodialysispatients. ASAIO J 48:565–569, 2002

22. Mapes DL, Lopes AA, Satayathum S, McCullough KP, Goodkin DA,Locatelli F, Fukuhara S, Young EW, Kurokawa K, Saito A, Bommer J,Wolfe RA, Held PJ, Port FK: Health-related quality of life as a

predictor of mortality and hospitalization: The Dialysis Outcomes andPractice Patterns Study (DOPPS). Kidney Int 64:339–349, 2003

23. Gibney N, Cerda J, Davenport A, Ramirez J, Singbartl K, Leblanc M,Ronco C: Volume management by renal replacement therapy in acutekidney injury. Int J Artif Organs 31:145–155, 2008

24. Shinzato T, Maeda K: Push ⁄ pull hemodiafiltration. Contrib Nephrol158:169–76, 2007

25. Roberts M, Lee DB-N: Wearable peritoneum-based system for continu-ous renal function replacement and other biomedical applications. USpatent 5944684, 8 ⁄ 31 ⁄ 1999

26. Ronco C, Fecondini L: The Vicenza wearable artificial kidney for peri-toneal dialysis (ViWAK PD). Blood Purif 25:383–388, 2007

27. Gura V, Beizai M, Ezon C, Polaschegg HD: The wearable artificial kid-ney (WAK). Contrib Nephrol 149:325–333, 2005

28. Gura V, Beizai M, Ezon C: CRRT for CHF; the wearable continuousultrafiltration system. ASAIO J 52:59–61, 2006

29. Constanzo MR: The role of ultrafiltration in the management of heartfailure. Congest Heart Fail. 14:19–24, 2008

30. Gura V, Ronco C, Nalesso F, Brendolan A, Beizai M, Ezon C,Davenport A, Rambod E: A wearable hemofilter: first human study ofslow continuous ambulatory ultrafiltration. Kidney Int 3:497–502, 2008

31. Davenport A, Gura V, Ronco C, Beizai M, Ezon C, Rambod E:A wearable hemodialysis device for patients with end-stage renal failure:a pilot study. Lancet 370:2005–10, 2007

32. Davenport A: Audit of the effect of dialysate sodium concentration oninter-dialytic weight gains and blood pressure control in chronic hemo-dialysis patients. Nephron Clin Pract 104:c120–125, 2006

33. Davenport A, Cox C, Thuraisingham R: The importance of dialysatesodium concentration in determining interdialytic weight gains inchronic hemodialysis patients: The PanThames Renal Audit. Int J ArtifOrgans 31:411–417, 2008

34. Lockridge Jr RS: Daily dialysis and long-term outcomes—The Lynch-burg Nephrology NHHD experience. Nephrol News Issues 13:16, 19,23–26, 1999

35. Lockridge Jr RS, Spencer M, Craft V, Pipkin M, Campbell D,McPhatter L, Albert J, Anderson H, Jennings F, Barger T: Nocturnalhome hemodialysis in North America. Adv Ren Replace Ther 8:250–256,2001

THE WEARABLE ARTIFICIAL KIDNEY 17