In Vitro and in Vivo Trials of Wearable Artificial

Kidney
Nikolay A. Bazaev, Victor M. Grinvald, Nikita M. Zhilo, Boris M. Putrya, Evgeniy V. Streltsov
National Research University of Electronic Technology
Moscow, Russia
bazaev-na@yandex.ru

Abstract—This paper presents results of in vitro and in vivo

trials of a wearable artificial kidney (WAK) based on peritoneal
dialysis. Materials and methods: in vitro trials were carried out
in a developed test bench, which imitates patient’s peritoneal
cavity and capable to measure main procedure parameters. In
vivo trial was carried out on a 15 kg dog in a healthy state to
evaluate the effect of WAK on blood biochemical indicators and
in an acute kidney injury state to test WAK’s blood purification
capabilities. Results: in vitro trials demonstrated that WAK
could eliminate urea, uric acid and creatinine from spent
dialysate with mean mass rates of 0.85±0.1 g/h, 0.10±0.04 g/h and
0.05±0.01 g/h respectively. Concentrations of Na+, Cl– and Ca2+
ions are kept in 10 % range from initial values. It is
demonstrated that WAK can operate without replacement of
expendable materials for 38 hours. During in vivo
trial (39,5 hours), WAK performed continuous peritoneal dialysis
with dialysis fluid regeneration. In acute kidney injury conditions
(34 hours) creatinine and uric acid were eliminated from
dialysing solution with the rate of 0,3 mg/h, urea was eliminated
at the rate of 0,15 g/h, total removed ultrafiltrate volume was
350 ml. At the end of the second stage blood biochemical
indicators stabilised in the range of normal values. Conclusion:
developed WAK can perform elimination of metabolites from
spent dialysis fluid with mass rates that are sufficient to maintain
stable, physiologically normal metabolite concentrations in
patient’s blood.
Keywords—continuous peritoneal dialysis with dialysate
regeneration; wearable artificial kidney; biomedical trial; dialysate
regeneration
I. INTRODUCTION
Patients with chronic renal failure have to maintain their
living organism by undergoing renal replacement therapy [1]:
hemodialysis (HD, extracorporeal blood purification via
metabolites’ diffusion through semipermeable dialyser
membrane (Fig. 1a); method is common, but patients have to
spend 4-6 hours near HD machine) or peritoneal dialysis (PD,
extracorporeal blood purification via metabolites’ diffusion
and osmotic transport across patient’s peritoneal membrane
(Fig. 1b) into dialysis fluid in peritoneum cavity; 0.5-2 l of
dialysis fluid is exchanged 3-4 times a day; method is less
widespread but is more physiological due to lower metabolites
elimination rate, and in most cases can be performed on an
ambulatory basis). The difference between HD and PD is
disclosed in Fig. 1.
Fig. 1. Dialysis available methods
During HD, blood is directly cleared in a dialyser in an
extracorporeal circuit. HD is carried out via HD machine, and
PD is a passive and more autonomous process of blood
purification. At present, several scientific groups work on a
prototype of a wearable artificial kidney (WAK) [2, 3],
because it has a number of advantages over hemodialysis:
– increasing patient’s mobility due to the reduction of its
size and weight;
– adjusting blood purification to human physiology
(metabolite elimination rate is closer to metabolite production
rate);
– reduction of water consumption (2 l of dialysate per one
procedure vs 120-150 l for hemodialysis) [4, 5];
– cost reduction (~€ 45 000 annually using WAK [6] vs
~€ 59 600 annually for continuous ambulatory peritoneal
dialysis [7]);
and the following advantages over continuous ambulatory
peritoneal dialysis (CAPD):
– lowering frequency of peritonitis due to rare dialysate
exchange (once a day);
– gaining life quality due to the ability to use WAK at
work or while travelling;
– reduction of water consumption (2 l of dialysate vs 8 l
for one day of CAPD).
An experimental prototype of WAK was designed,
assembled and tested. It carries out peritoneal dialysis with
regeneration of spent dialysate in its extracorporeal circuit.
Results of in vitro trials of WAK are disclosed in this paper.

1877 978-1-5386-4340-2/18/$31.00©2018 IEEE

 II. MATERIALS AND METHODS
WAK consists of a hydraulic circuit, performing
recirculation and regeneration of dialysate, and an electronic
circuit that performs device controlling (functional diagram of
a biotechnical system with developed WAK is presented in
Fig. 2 in the top). Two roller pumps (Thomas SR10/50)
recirculate dialysis fluid through hemofilter (Baxter Aquamax
HF03): the first one from peritoneum cavity through blood
ports (intracorporeal circuit); the second one from dialysate
port through regeneration unit (extracorporeal circuit).
Regeneration of waste dialysis fluid involves elimination
of metabolism products such as urea, creatinine and uric acid.
This is achieved by combining electrochemical (for urea
elimination) and sorption (for the elimination of larger solutes)
methods. Dialysis regeneration unit consists of the
electrolyser, two sorption elements (before and after
electrolyser), degasser and a pump with correction fluid
(optional, was not used during experiments). Sorption element
represents a container with an inlet and outlet sockets that hold
55 g of FAS activated carbon (sorption capacity for creatinine
130 umol/g, granule size ~2 mm). Sorption elements can be
used starting from the very beginning of the PD.
Electrolyser represents a chamber with input and output
sockets for dialysis fluid that flows between
24 electrodes (surface area 50 cm2 each, 1200 cm2 total, 2 mm
electrode spacing). Electrodes represent working section that
interacts with dialysis fluid and leads for electrical contacts
Fig. 2. Device functional diagram (top) and test bench functional
diagram (middle and bottom): UF – ultrafiltrate, 1 – WAK with tubing
set, 2 – infusion pump, 3 – ion and biochemical analysers, 4 – thermostat
with bulb, 5 – laptop with running WAK service software
1878
from outside of the electrolyser. Working section of the
electrode is a square titanium support with a platinum coating
on both sides. Apertures in the cover plate of the electrolyser
are for electrical contacts (for electrode’s leads). Outside parts
of the leads are hydraulically isolated from the inner
compartment of the electrolyser.
Developed WAK is enclosed in a backpack; its overall
weight is 3.5 kg. WAK is controlled by a smartphone via
Bluetooth connection or by laptop. The electronic circuit
monitors fluid pressure, temperature and conductivity and is
able to alarm user of probable errors. Expendable materials for
WAK are tubing set, two sorption elements and electrolyser.
A. In vitro trials
In order to carry out WAK trials, a test bench was
developed. Its functional diagram and visual appearance are
presented in fig. 2 at the bottom. The test bench consists of:
– patient imitator: airtight bulb which imitates peritoneal
cavity of a patient, placed in thermostat, which keeps bulb at
37 °C, and connected to tubing set and infusion pump, which
imitates metabolite production in a patient. Patient imitator
works with preliminary set parameters (temperature, flow rate,
metabolites addition rate);
– the tubing set with outputs for connection to the WAK
and probe sampling;
– measuring equipment for evaluation of the chemical
composition of dialysis fluid: biochemical analyser and
electrolyte analyser;
– other measuring equipment: atmospheric moisture and
temperature meter, stopwatch, analytical scales, set of
measuring flasks, laptop, power unit, set of supports and
syringes.
The model fluid of waste dialysate solution was the
combination of the solution for peritoneal dialysis
CAPD/DPCA 2 and metabolites: urea, creatinine and uric acid
with concentrations of 36,3±2.2 mmol/l, 900±70 umol/l and
265±85 umol/l. pH level of the model fluid was 5…6.
Mentioned metabolites were chosen because they are common
markers of dialysate efficiency and must be eliminated by
WAK. Initial concentrations of metabolites were at the top
level of their concentrations at the end of peritoneal
equilibration test (PET).
Dialysate flow rate in both extracorporeal and
intracorporeal circuits was in the range of 40…70 ml/min. To
imitate metabolites production infusion pump gradually added
during every hour 2 ml of water with 0.8 g of urea, 0.1 g of
creatinine and 0.05 g of uric acid into the patient imitator.
Dialysate samples were taken from the bulb that is why results
represented the tendency of mean values.
During the experiment, WAK operated in automatic mode,
except electrolyser. It was manually set to a galvanostatic
mode (constant electrical current), and in the first 14 hours
current density was 4.5 mA/cm2, and 6.0 mA/cm2 during next
24 hours.
 Expendable materials were not changed during the 0.85±0.1 g/h. We can see that by changing the current density
experiment. To gather enough statistical data experiment was we can change the elimination rate with no interference to the
repeated 15 times. pH of dialysate, but as Fig. 3 shows it can be correlated with
Cl– production because free chlorine release increases with
B. In vivo trials decreasing of urea concentration [8].
A dog with 15 kg body mass was chosen as a laboratory
animal. Selection criteria: a healthy animal with a sufficient Creatinine and uric acid were eliminated by sorption
body weight and, as a consequence, with a large volume of the elements. Experiments show that concentration of these
abdominal cavity for effective PD. Two abdominal catheters substances can be kept on the level of ½ of the initial
connected to the extracorporeal contour of the apparatus were concentration of dialysis fluid. In other words, sorption
installed in the abdominal cavity. elements always bind half of the metabolites that present in
dialysis solution. In comparison to continuous ambulatory
Research methods: blood sampling and evaluation of peritoneal dialysis, such system can theoretically raise
laboratory indicators (creatinine, electrolytes, total protein clearances of creatinine and uric acid up to 10 times. Average
with fractions) using a blood analyser; sampling of the spent creatinine and uric acid elimination rate during the experiment
peritoneal dialysis solution and measurement of the main were 0.10±0.04 g/h and 0.05±0.01 g/h respectively.
dialysis markers’ concentrations (urea, creatinine, uric acid)
using a biochemical analyser and ion concentrations (Na+, K+,
Cl–, Ca2+) using an electrolyte analyser.
The apparatus was tested in two stages:
1) a laboratory animal with healthy kidneys (to assess the
effect of the apparatus solely on the biochemical parameters of
the blood) – the first stage of the experiment lasting 8 hours.
In the beginning, the peritoneal cavity was filled with a
peritoneal dialysis solution Balance 1.5 % (0.6 l in the cavity,
0.4 l in the WAK), after 2 hours the device was switched on
without electrolysis, then after 5 hours, electrolysis was turned
on for 30 minutes;
2) a laboratory animal with acute renal failure (to assess
the effectiveness of the device in terms of artificial
purification of the body). In order to cause acute renal failure,
a laboratory animal was administered 200 ml of radiocontrast
substance, which causes obstruction of the renal tubules and
temporary acute renal failure. Continuous cyclic purification
of the solution for peritoneal dialysis was carried out, with
periodic activation of an electrolyzer for the elimination of
urea. Two solutions were used: Extraneal and Balance 1.5 %
(0.6 l in the cavity, 0.4 l in the WAK).
During the second stage of the trial, an increase in the
creatinine concentration in the blood to 162 umol/l was
recorded. In addition, throughout the second stage of the test,
the laboratory animal had anuria. This allows us to make a
conclusion about the onset of acute kidney injury in a
laboratory animal during the second stage of the test.
III. RESULTS AND DISCUSSION
A. In vitro trials
Experiment results are presented in Fig. 3. It can be seen
that pH of the dialysis fluid tends to 7.0. That was
demonstrated on several different dialysis fluids with initial
pH in the range from 5.2 to 7.0. That is one of the most
interesting results because usually electrolysis acidizes the
dialysis fluid and addition of buffer fluid is needed. However,
a very promising combination of electrolysis with activated
carbon was found that normalises pH of dialysate. As to the
urea elimination, during first 14 hours, its average value was
0.68±0.07 g/h. Then current density was raised up to Fig. 3. Concentrations of metabolites, ions and pH dynamics in the
6.0 mA/cm2 and the average urea elimination rate was dialysate

1879
 It can be seen that WAK can eliminate urea from dialysate
with the mean mass rate of 0.85±0.1 g/h (when electrode
current density is 6 mA/cm2), creatinine and uric acid with
mean mass rates of 0.10±0.04 g/h and 0.05±0.01 g/h
respectively. WAK is capable of performing its function
without replacement of expendable materials for 38 hours.
This period can be extended by adding correction fluid that
contains Ca2+, because its concentration lowers during the
procedure from 1.18 mmol/l to 0.86 mmol/l. At the same time
concentration of K+ ions slowly raises from 0.065 mmol/l to
0.200 mmol/l. Concentrations of Na+ and Cl– ions are quite
stable: Na+ from 131.97 mmol/l at the start to 132.60 mmol/l
at the end of experiment with maximum of 133.20 mmol/l and
minimum of 125.40 mmol/l; Cl– from 101.25 mmol/l at the
start to 109.50 mmol/l at the end of experiment with maximum
of 109.50 mmol/l and minimum of 96.80 mmol/l. The impact
of ultrafiltration on the concentration dynamics and evaluation
of WAK was not taken into consideration. However, at present
stage of work, the ability to eliminate metabolites with WAK
was shown. Developed WAK is able to take out excess fluid
from the peritoneum cavity.
B. In vivo trials
The results of the tests on the model of a healthy organism
are presented in Fig. 4 and in Table 1.
Fig. 4 shows the dynamics of urea in a solution for
peritoneal dialysis. It can be seen that the sorbent did not
eliminate urea during the first 6 hours and 45 minutes; it was
accumulated in the solution. After switching on the
electrolysis, urea concentration decreased by 2.54 mmol/l (the
mass elimination rate was 150 mg/h) in an hour. At the same
time, the sorbent eliminated creatinine and uric acid. In the
first three hours, regeneration of the solution for peritoneal
dialysis was not performed, as a result of which the
concentration of metabolites in the solution for peritoneal
dialysis increased. After switching on the device, there was a
decrease in the concentration of creatinine and uric acid, and
their stabilization at 35 mmol/l. Analysis of the concentration
of ions showed their insignificant changes during the test, the
pH of the solution shifted to the alkaline region.
Fig. 4. Concentrations of metabolites, ions and pH dynamics in the
dialysate in healthy animal study
1880
TABLE I. BLOOD BIOCHEMICAL ANALYSIS OF HEALTHY BIOLOGICAL
OBJECT (REFERENCE VALUES FOR DOGS ARE IN BRACKETS)
Time
Value
00:00 7:40
Ʉ+, mmol/l (4,3-6,2) 4.4 4.8
Na+, mmol/l (138,0-164,0) 149.0 158.6
Total protein, g/l (40,0-73,0) 70.0 63.0
Creatinine, umol/l (26,0-120,0) 96.0 74.0
During the experiment, blood biochemical parameters
changed insignificantly but remained in the range of normal
values. Thus, it can be said that WAK did not worsen the
condition of the animal for 5 hours, and allowed to improve
the efficiency of the peritoneal dialysis. According to the
Figure 3, it is possible to estimate the amount of creatinine and
uric acid removed during the standard peritoneal dialysis
procedure – 75 umol and 55 umol, respectively. While using
the WAK, it is possible to remove 145 umol and 85 umol of
creatinine and uric acid for the same 6 hours, respectively.
The results of tests on the animal model with acute renal
failure are presented in Fig. 5 and Table 2.
Fig. 5 shows the dynamics of urea concentration in a
peritoneal dialysis solution. It should be noted that the
activation of the electrolyser led to a sharp decrease in the
concentration of urea. In addition, worthy of attention is that at
the end of electrolysis at 27 hours of the experiment, the urea
concentration stabilised at a level of 3 mmol/l, which
corresponds to the maximum urea concentration in the
experiment with healthy kidneys and may indicate restoration
of kidney function. During the second stage of the test, the
average urea removal rate was about 50 mg/h, while the WAK
allowed removing urea at a mass rate of 150 mg/h. For
comparison, in dogs with comparable body weight, healthy
kidneys remove urea at a rate of about 250 mg/h.
Fig. 5. Concentrations of metabolites, ions and pH dynamics in the
dialysate in animal with acute kidney injury
 TABLE II. BLOOD BIOCHEMICAL ANALYSIS OF BIOLOGICAL OBJECT
WITH ACUTE KIDNEY FAILURE (REFERENCE VALUES FOR DOGS ARE IN
BRACKETS)
Time
Value
09:30 17:30 33:30
Ʉ+, mmol/l (4,3-6,2) 3.9 3.9 4.1
Na+, mmol/l (138,0-164,0) 149.0 150.4 146.0
Total protein, g/l (40,0-73,0) 59.0 60.0 45.0
Creatinine, umol/l (26,0-120,0) 129.0 162.0 44.0
Figure 5 shows the concentration dynamics of uric acid
and creatinine. Stabilisation of their concentrations in the
range of 25…35 umol/l with temporary jumps to high
concentrations 30…50 umol/l can be seen. Analysis of the
creatinine and uric acid concentrations before and after the
regeneration block allows us to tell that they are cut in half
after a single passage through it. Thus, during the experiment,
110 mg of creatinine and 190 mg of uric acid were removed.
The ionic composition of peritoneal dialysis solutions for
was in the range of ± 10 % of the initial value, the pH of the
solution stabilized near normal blood pH values, without
jumps to the alkaline or acidic regions. The biochemical blood
test also confirms the stable state of the laboratory animal
throughout the test period.
During the experiment, 350 ml of ultrafiltrate were
removed on medical indications.
Acute renal failure was induced in the dog, which was
confirmed by biochemical blood tests and the metabolites
concentrations dynamics in the dialysis solution. After 34
hours of dialysis, the biochemical blood values were
normalized with a WAK and the experiment was stopped.
IV. CONCLUSION
In vitro trials of wearable artificial kidney experimental
prototype was carried out on the test bench. WAK
demonstrated the capability of metabolic waste elimination
from spent dialysis fluid with mass rates that are sufficient to
maintain stable, physiologically normal metabolite
concentrations in patient’s blood. It is shown that expendable
materials can be replaced at least once in one and a half days,
which lowers risk of peritonitis up to 6 times, increases
ergonomics of WAK and patient’s quality of life.
In vivo trials of the WAK on the animal model were
carried out. Tests on a healthy animal showed that the
1881
apparatus does not cause a pathological effect on the
biochemical parameters of the blood; it makes it possible to
eliminate metabolites from the solution for peritoneal dialysis.
During the tests on the animal with acute kidney injury, the
following indicators of dialysate purification were achieved:
the mass removal rate of urea from the spent peritoneal
dialysis solution 0.15 g/h, the mass removal rate of creatinine
and uric acid 0.3 mg/h. It is expected that the effectiveness of
blood purification in patients with chronic renal failure will be
2-4 times higher. This is due to the larger area of the
peritoneal membrane in human and a larger volume of
solution for peritoneal dialysis in the peritoneal cavity.
Nevertheless, despite the good results obtained during the
experiment, the need for additional studies should be noted. In
particular, it is necessary to conduct additional in-depth
studies of the electrolysis of the spent dialysate solution,
continue to search for materials for the preparation of
electrodes for the replacement of precious metals and to
continue to study the possibilities of even more stable pH and
ionic solution composition for peritoneal dialysis.
ACKNOWLEDGMENTS
This work was partly financed by the Ministry of
Education and Science of the Russian Federation (project ID
14.578.21.0011, 05.06.2014).
REFERENCES
[1] Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW. National
Kidney Foundation practice guidelines for chronic kidney disease:
evaluation, classification, and stratification. Annals of internal medicine.
2003;139(2):137-147.
[2] Armignacco P, Garzotto F, Neri M, Lorenzin A. WAK Engineering
Evolution. Blood purification. 2015;39(1-3):110-114.
[3] Ronco C, Fecondini L. The Vicenza Wearable Artificial Kidney for
Peritoneal Dialysis (ViWAK PD). Blood purification. 2007;25(4):383-
388.
[4] Blagg CR. Home haemodialysis: ‘home, home, sweet, sweet
home!’(Review Article). Nephrology. 2005;10(3):206-214.
[5] Gura V, Macy A, Beizai M, Ezon C, Golper TA. Technical
Breakthroughs in the Wearable Artificial Kidney (WAK). Clinical
Journal of the American Society of Nephrology. 2009;4(9):1441-1448.
[6] Nanodialysis. Nanodialysis Miniature Dialysis System brochure.
http://www.nanodialysis.nl/media/Nanodialysis_Miniature_Dialysis_Sys
tem_Brochure_2015.pdf. Accessed March 27, 2017.
[7] Eriksson JK, Neovius M, Jacobson SH, Elinder CG, Hylander B.
Healthcare costs in chronic kidney disease and renal replacement
therapy: a population-based cohort study in Sweden. BMJ open.
2016;6(10), e012062.