000184116

Nephron 45: 196 201 (1987)
& 1987 S. Karger AG. Hasel
0028-2766/87/0453 0I96S2.75/0
Protein Losses in Patients Receiving Continuous Ambulatory Peritoneal Dialysis
G.A. Young, A.M. Brownjohn, F.M. Parsons
Renal Research Unit, General Infirmary, Leeds, UK
KeyWords. Continuous ambulatory peritoneal dialysis ■Blood proteins • Albumin • Prealbumin • Transferrin ■
Globulins • Glycoproteins • Haptoglobin • Peritoneal dialysis
Abstract. Total protein and 12 specific proteins were measured in dialysates from 8 patients on continuous
ambulatory peritoneal dialysis during training. Mean daily loss of total protein was 10.5 g and this included 5.2 g
albumin, 805 mg of the immunoglobulins G, A and M, 323 mg transferrin and 530 mg of the remaining 7 proteins
measured. The plasma to dialysate ratio of protein concentrations correlated with the natural logarithm of molecular
weight, suggesting that proteins in dialysate are an ultrafiltrate of plasma. A greater loss of proteins overnight was due
to longer dwell time as the mean rate of loss was similar for all exchanges. Losses were similar with 1.36% and 3.86%
dextrose fluids, suggesting that the initial effects of hypertonicity are diminished or reversed by dilution and
absorption of dextrose. Daily outflow volumes for 4 patients correlated inversely with the quantities of several
proteins removed, probably due to effects of osmolality. It is concluded that protein losses are related to plasma
concentration, molecular weight and osmolality of the dialysis solution and to the physiology of the patient.
Continuous ambulatory peritoneal dialysis (CAPD) is
a valuable alternative to haemodialysis for the treatment
of end-stage renal disease. The loss of proteins into the
dialysis fluid was first measured during intermittent peri­
toneal dialysis (IPD) [1] and has since been studied clin­
ically and experimentally. Comparisons between IPD
and CAPD have shown similar losses of 30-100 g/week
[2], Recently there have been several investigations into
the causes and consequences of protein losses during
peritoneal dialysis and these have been reviewed [3] but
very few detailed studies of specific proteins have been
made. In this study we have evaluated the losses of total
protein and 12 specific proteins into dialysis fluid during
CAPD and have identified some of the factors contribut­
ing to these losses.
Patients and Methods
Patients
Eight patients, 4 females and 4 males, aged 28-58 years (table 1)
were studied during their initial 10- to 13- day training period whilst
in hospital. All patients had an OreopoulosorTenckhoff peritoneal
catheter and had received intermittent peritoneal dialysis for be­
tween 5 days and 22 months. They were dialysed continuously with
4 exchanges daily using 2 I plastic bags of Travenol® Dianeal 137
except for 1 small patient (No. I) who was prescribed 1 I bags.
Dextrose (anhydrous) concentration was usually 1.36 g/100 ml (356
exchanges) although a 3.86 g/100 ml dehydrating dialysis solution
was used intermittently in 4 patients to maintain constant daily
weight (23 exchanges). Mean dietary protein intake was 1.4 g/kg of
body weight and calories 27 Kcal/kg in addition to calories derived
from dextrose in the dialysis fluid. Vitamin supplements and alumi­
nium hydroxide were prescribed for most patients.
Methods
Body fat and arm muscle circumference (AMC) were calculated
from anthropometric measurements [4] and body surface area was
calculated from height and weight [5). Osmolality of dialysate was
measured using an Osmometer (Fisons). Total protein in dialysate
and plasma was measured using a Beckmann® Trace I analyser and
specific proteins measured using a Beckmann® Immunochemical
analyser or by radial immunodiffusion. The recovery of proteins
from 1.36 and 3.86% dextrose dialysis fluid containing added plasma
was between 95 and 104% except for immunoglobulin M (85%), p2
microglobulin and caeruloplasmin (75%). All proteins were measu­
rable in dialysis fluid without any concentration. Heparinised blood
samples were collected and the plasma separated and stored at -70
°C. During the study all dialysates were collected at the completion
of each exchange, mixed and measured. Aliquots from each ex-
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
Introduction
Protein Losses during CAPD
197
Table I. Sex, age. previous l PD and anthropometric data for 8 patients at the start of CAPD
Patient
Sex
Age
years
Previous
treatment on
IPD, months
Height
cm
Weight
kg
Body surface
area, m2
Fat
kg
AMC
cm
1
2
3
4
5
6
7
8
F
F
F
F
M
M
M
M
28
49
57
58
38
51
52
55
< 1
<1
22
3
18
<1
12
<1
142
164
155
160
168
161
185
163
30.4
63.5
57.0
65.4
78.7
52.0
80.0
60.0
1.1
1.7
1.6
1.7
1.9
1.5
2.0
1.6
5.1
18.5
20.2
22.3
19.5
8.6
26.2
16.5
14.9
24.3
21.6
24.6
29.6
23.4
28.0
26.3
162 ±12
60.9 ±15.7
1.64 ±0.27
17.1 ±7.0
24.1 ±4.5
48.5 ±10.4
M ean±SD
Table II. Losses of total protein and 6 specific proteins in CAPD dialysate during training; plasma protein concentrations at the start and
end of the study are also shown
1
2
3
4
5
6
7
8
M ean±SD
Study
days
10
10
13
13
13
13
10
13
Start of CAPD
End of study
Outflow volume
l/day
4.4 ±0.5
9.1 ±0.1
8.5 ±0.5
8.2 ±0.6
9.0 ±0.3
9.4 ±0.7
9.0±0.6
8.3 ±0.2
8.2 ±1.6
Dialysate protein losses
total protein
g/day
albumin
g/day
transferrin
mg/day
IgG
mg/day
IgA
mg/day
IgM
mg/day
8.0 ± 2.1
9.8 ±1.0
10.7 ±3.6
10.9 ± 1.8
11.4 ± 1.6
11.6 ±2.3
10.7 ± 1.5
11.0 ± 1.4
10.5 ± 1.2
Plasma protein
4.8 ±1.5
5.4 ±0.5
4.7 ± 1.8
5.2 ±1.2
5.3 ±1.4
4.5 ± 1.1
5.5 ± 1.7
5.9 ±0.6
5.2 ±0.5
concentrations
299 ±80
323 ±4.6
325 ±116
389 ±78
361 ±93
267 ±56
249 ±67
373 ±58
323 ±5.0
718 ±299
684 ±91
572 ± 312
675 ±177
489 ±17
672±238
596± 162
579 ±68
623 ±77
106 ±41
130 ± 13
99 ±49
132 ±42
207 ± 51
91 ±28
82 ±28
147 ± 15
124 ± 40
42 ±21
51 ±21
51 ±23
40 ± 7
103 ±45
50 ±23
48 ± 14
76± 16
58 ± 21
g/dl
g/dl
mg/dl
mg/dl
mg/dl
mg/dl
3.9 ±0.4
3.7 ±0.4
264 ±40
269 ±47
914 ±212
864 ±205
253 ±54
221 ±54
192 ±45
184 ±27
6.9 ±0.8
7.0 ±0.7
change and the pooled samples containing all the days’ exchanges in
proportion to their volumes, were stored in containers without
preservative at 20 °C. The Student t test was used for comparison of
both paired and unpaired values.
Results
Palien I Data:
Table I shows the sex, age, previous dialysis treatment,
height, weight, calculated body surface area, fat and
AMC for the 8 patients. There were no significant differ­
ences between the sexes for age, height, weight, body
suface area, fat or AMC. No peritonitis or other illness
occurred during the study. All the patients had previously
been on IPD; 4 for periods of 3-22 months and 4 for less
than 1 month. Patient 4 alone had had a single attack of
peritonitis 2 months before the start of CAPD.
Protein Losses in Dialysates:
Mean daily losses of total protein and 12 specific
proteins into the dialysate during the training period for
each of the 8 patients are shown in tables II and III. The
mean daily loss and mean initial and final plasma con­
centrations for each protein in the 8 patients are also
shown. The mean loss of total protein was 10.5 g and
included 5.2 g albumin (49%), 805 mg immunoglobulins
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
Patient
Young/Brownjohn/Parsons
198
Table III. Losses of 7 specific proteins in CAPDdialysale during training: plasma protein concentrations at the start and end of the study
are also shown
Patient
1
2
3
4
5
6
7
8
M ean±SD
Start of CAPD
End of study
Study
days
10
10
13
13
13
13
10
13
Dialysate protein losses, mg/day
prealbumin
AAG
haptoglobin
AMG
caerulo­
plasmin
(i2 micro­
globulin
C3
32 ±12
40 ± 13
46 ± 10
46 ± 11
55± 16
64 ± 19
41 ± 7
44 ± 9
46 ± 10
Plasma protein
97 + 26
95 ±11
167 ± 66
119 ±20
176 ±40
204 ± 91
158 ±29
222 ±25
155 ±47
80 ±30
44 ± 7
121 ±78
52 ± 19
35 ± 10
160 ±39
62 ±18
156 ±22
89 ±50
87 ±39
67 ±12
127 ± 102
47 ± 17
67 ±21
73 ±22
51 ±18
95 ±20
77 ±26
19 ± 4
26 ±5
34± 16
23 ±6
29 ± 11
34 ± 12
24 ±10
41 ±6
29 ±7
17 ± 2.8
12 ± 1.9
31 ±3.7
29 ±3.8
24 ±2.9
21 ±3.5
25 ±2.7
24 ±2.9
23 ± 6
76 ±24
97 ± 18
120 ±65
85 ±20
169 ± 65
150 ±35
120 ± 31
124 ± 14
115 ±32
concentrations, mg/dl
39 ±4.0
39 + 4.0
109 ± 31
107 ±36
249 ±74
256± 74
54 ± 13
47 ±10
2.6 ±0.7
2.7 ±0.6
155 ±32
166 ±31
2 10 ± 116
207 ±108
this low-molecular-weight protein were 27.3 mg/day and
18.5 mg/day (p< 0.025).
Plasma Concentrations.The mean loss for each protein
for all the patients correlated with their mean concentra­
tions in plasma at the start of training (r = 0.99: p < 0.001),
suggesting that the proportions are similar in dialysate
and plasma. The wide ranges of concentrations contri­
buted to this high correlation coefficent and when the
highest concentrations for albumin and lgG are excluded
then we found r = 0.59; p<0.05.
Molecular Weight. The concentrations of the proteins
in plasma are ranked in figure I and the corresponding
Factors Contributing to Protein Losses
amounts in dialysate and the molecular weights are
Sex. Differences in mean protein loss between the shown. Several proteins in dialysate are in smaller
sexes were not significant except for a t acid glycoprotein amounts than expected from plasma concentrations and
(p<0.02) and complement C3 (p<0.02). The use of 1 1 this can be attributed to higher molecular weights, e.g., a2
exchanges and the smaller body size of the female patient macroglobulin, IgM and caeruloplasmin. In contrast,
1may have contributed to the smaller mean loss for these larger amounts in dialysate are associated with lower
two proteins.
molecular weights, e.g., albumin, transferrin, a, acid gly­
Age and Anthropometry. The losses for each of the 12 coprotein and p2 microglobulin. This relationship is
specific proteins were unrelated to age ot to anthropo­ shown in figure 2. The plasma :dialysate ratio of protein
metric measurements of the patients although total pro­ concentrations correlated with the natural logarithm of
tein did correlate with AMC (r = 0.78; p<0.05).
molecular weight (r = 0.92; p<0.001). This suggests that
Previous Intermittent Peritoneal Dialysis. There was no proteins in dialysate are an ultrafiltrate of plasma.
significant difference between protein losses in the 4
Losses during Each Exchange. Losses of 3 proteins,
patients who had previously been on I PD for the longest albumin, transferrin and haptoglobin were measured for
periods of 3-22 months and those for less than I month the 8 patients during 270 daytime exchanges and 83
except for p2 microglobulin. The comparative losses of overnight exchanges using 1.36% dextrose. The mean
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
G, A and M (8%), 323 mg transferrin (3%) and 530 mg of
the remaining 7 proteins (5%). The daily loss for each
protein was equivalent to 31 142 ml plasma, i.e., a clear­
ance of 0.02-0.10 ml/min [loss per day or per minute
divided by plasma concentration (g/ml)] with the excep­
tion ofp2 microglobulin which was equivalent to 938 ml
plasma per day or 0.65 ml/min. There was no significant
change in the daily loss or the concentration in plasma for
each protein during the training period. Over the 13 days
the coefficient of variation for the mean daily total pro­
tein loss was 18%.
Protein Losses during CAPD
199
Molecular weight, In
Fig. 1. Initial plasma concentrations for each of 12 proteins are
ranked and the corresponding amounts in dialysate and their molec­
ular weights are shown. Alb = Albumin; IgG = Immunoglobulin G;
Trf = transferrin; IgA = immunoglobulin A; AMG = a : macroglob­
ulin; Hapt = haptoglobin; IgM = immunoglobulin M; C3 = complement C3; AAG = ct| acid glycoprotein: Caer = caeruloplasmin:
PreA = prealbumin; [LMG = microglobulin.
rates of protein loss during daytime (±SD ) exchanges
were: albumin 218± 60 mg/h, transferrin 13.7± 4.0 and
haptoglobin 4.2 ±3.0 mg/h and those at night time were
not significantly different (unpaired t test). The total
losses at night time were greater because the exchanges
were longer, i.e., 9.5 compared with 4.8. The first 2 day­
time exchanges at the start of training were not included
in the calculation of daily losses because of residual
protein and dialysate in the peritoneum from previous
IPD.
The Effects of Osmolality o f Dialysis Fluid on Protein
Loss
During a total of 377 exchanges in the 8 patients,
3.86% dextrose was required for only 24 exchanges in 4
patients. The mean rates of loss for albumin, transferrin
and haptoglobin were not significantly different using
either 1.36 or 3.86% dextrose during the overnight or
daytime exchanges. The effects of the movement of water
into the peritoneum during CAPD on the amounts of
protein removed were evident in 4 patients maintained on
1.36% dextrose fluid. The daily losses of several proteins
were inversely correlated with the volumes of dialysate
removed. In patient 1an inverse correlation was observed
between volumes of fluid removed and total protein
(r = 0.75; p<0.05), albumin (r = 0.72; p<0.05) as shown
in figure 3 and also with caeruloplasmin (r = 0.70;
p < 0.05). Similar correlations were observed in patient 6
for transferrin (r = 0.79; p<0.01), caeruloplasmin
(r = 0.56; p<0.05) and haptoglobin (r = 0.73; p <0.01),
and for albumin and haptoglobin respectively in patients
2 and 5. The volumes removed also correlated with the
osmolality of the dialysate in patient 1. This suggests, that
the accumulation of excess fluid in the peritoneum de­
creased the osmolality and may have contributed to a
reduced loss of proteins. Outflow volume was unrelated
to protein loss in the other 4 patients.
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
Fig. 2. Plasma: dialysate ratio correlated with the natural loga­
rithm of molecular weight. Abbrevations as in figure 1.
200
Young/ Brownjohn/Parsons
(820,000) are retarded and low-molecular-weight pro­
teins such as ¡)2microglobulin (11,820) are increased. This
is characteristic of the ultrafiltration process that occurs
across the other body fluid barriers such as plasma to
cerebrospinal fluid [6], synovial fluid [7], amniotic fluid
[8] and also ascitic fluid [9]. Some effects of molecular
weight on protein loss during CAPD have been suggested
in a study of 6 proteins [10] and from a review of previous
data [3]. The quantity of |)2microglobulin in dialysate is
very high and the possibility of synthesis by macrophages
within the peritoneum, as previously suggested for col­
lagen [11] and glutathione peroxidase [12], cannot be ex­
cluded.
Effects o f Physiology on Protein Loss
The amount of protein removed during CAPD is re­
Fig. 3. Inverse correlations between losses of total protein,
lated to certain anatomical and physiological features of
albumin and osmolality against outflow fluid volume in patient I.
the patient [3]. In a previous study the total protein in
dialysate correlated with body surface area and this was
equated with anatomical surface area [15], We did not
Discussion
observe this relationship but protein loss correlated with
arm muscle circumference. Smaller protein losses in
Protein Losses relative to Plasma Concentration and
women may be associated with the smaller peritoneal
Molecular Weight
surface area although such differences between men and
The daily losses of total protein, albumin, transferrin, women were confined to ct| acid glycoprotein and com­
immunoglobulins G, A and M, and complement C3 were plement C3 and other causative factors cannot be ex­
of a similar order to that found by Blumenkrantz et al. [2] cluded. The comparability of losses between patients
although comparable measurements for prealbumin, a, suggest that these were unaffected by any clinical differ­
acid glycoprotein, haptoglobulin, a 2 macroglobulin, ca­ ences or by previous treatment with I PD. (12 Microglobueruloplasmin and IT microglobulin are not available. The lin was the only protein lost in greater amounts in those
concentration in dialysate was related not only to plasma patients previously maintained on 1PD for longer peri­
concentration but also to molecular weight. Our results ods, suggesting that any increase in porosity of the perito­
suggest that loss of higher-molecular-weight proteins neum to proteins was limited to those of low molecular
such as immunoglobulin (950,000) and a2macroglobulin weight. None of the patients had peritonitis or other
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
The Effects o f Dialysis Fluid on Protein Loss
The effects of osmolality were evident only in patients
maintained solely on a 1.36% dextrose solution where the
amounts of proteins removed were inversely related to
the outflow volumes of the dialysate. This suggests that
the transfer of excess fluid into the peritoneum may
decrease osmotic pressure and reduce protein loss. Des­
pite this relationship with osmolality the use of hyperos­
molar dextrose solution, both in this and several previous
studies, did not cause a greater loss of protein [2, 13].
Clearly any effects of osmolality are small and if in­
creased protein losses occur during hyperosmolar dialy­
sis over the first 3 h [14] they could subsequently decrease
with the fall in osmotic pressure caused by the absorption
of dextrose and expansion of the fluid volume.
Protein LossesduringCAPD
8
9
Consequences o f Protein Loss
The consequences of protein losses during CAPD
have not been fully elucidated. During this study there
was no decrease in the plasma concentrations of proteins
measured and in previous short-term studies nutritional
status and albumin haemostasis were maintained in wellnourished patients [17]. During long-term studies a de­
crease in plasma proteins, body muscle and total body
nitrogen may occur [18,19], particularly when the dietary
protein intake is less than 1.2 g/kg body weight [19], Any
depletion of plasma proteins will inhibit many vital func­
tions, particularly transport mechanisms and response to
infection [19]. Replenishment of specific proteins will
also depend on their rates of biosynthesis. Consequently,
if the daily loss equals or exceeds the amount synthesised,
depletion will occur. This is most likely for low-molecu­
lar weight compounds where the clearance is high, e.g., pi
microglobulin. Further studies of the loss of small pro­
teins are required.
Acknowledgements
We wish to thank Mrs Shirley Hobson, Mr Albert Taylor and
Miss Mandy Thompson for their invaluable help. We are grateful to
the nursing staff of the General Infirmary at Leeds and the Leeds
Area Health Authority who kindly supported this work.
10
11
12
13
14
15
16
17
18
References
1 Berlyne, G.M.: Jones, J.H.; Hewitt, V.: Nilwarangkur, S.: Pro­
tein loss in peritoneal dialysis. Lancet, i: 738-741 (1964).
2 Blumenkrantz, M.J.: Gahl, G.M.; Kopple, J.D.; Kamdar. A.V.;
Jones, M.R.; Kessel, M.; Coburn, J.W.: Protein losses during
peritoneal dialysis. Kidney int. 19:593-602 (1981).
3 Dulaney, J.T.; Hatch, F.E.: Peritoneal dialysis and loss of pro­
teins: a review. Kidney int. 26:253-262 (1984).
4 Durnin, J.V.G.A.: Womersley, J.: Body fat assessed from total
body density and its estimation from shinfold thickness: mea­
surements on 481 men and women aged from 16 to 72 years. Br. J.
Nutr. 32:77-97 (1974).
5 Sendroy, J.; Cecchini, L.P. Determination of human body sur­
face area from height and weight. J. appl. Physiol. 7:1-12 (1954).
6 Felgenhauer. K.: Protein size and cerebrospinal fluid composi­
tion. Klin. Wschr. 52:1158-1164(1974).
7 Levick, J.R.: Permeability of rheumatoid and normal human
19
synovium to specific plasma proteins. Arthritis Rheum. 24:
1550-1560(1981).
Burnett, D.; Wood, S.M.: Studd, J.W.W.; Lewis, B.V.: Bradwell.
A.R.: A study of the relationship between maternal serum and
amniotic fluid proteins in early and late pregnancy. Protides biol
Fluids 24:161-164 (1976).
Henderson, J.M.; Stein, S.F.: Kulner, M.: Wiles. M.-B.: Ansley,
J.D.; Rudman. D.: Analysis of twenty three plasma proteins in
ascites. The depletion of fibrinogen and plasminogen. Ann.
Surg. 192:738-742 (1980).
Giangrande, A.: Limido. A.tCanto. P.; Allaria. P.: SDS-Polyac­
rylamide electrophoresis of protein loss during continuous am­
bulatory peritoneal dialysis. Clin.Lab. 10:117 12 (1980).
Kowalewski, J.: Tomaszenki, J.: Hanzlik, J.; Zawislak. H.:Zbikowska. A.: The elimination of free, peptide-bound and proteinbound hydroxyproline into dialysate during peritoneal dialysis
in patients with renal failure. Clin. Chim. Acta. 34: 123-126
(1971).
Young, G.A.; Brownjohn, A.M.; Turney, J.H.: Glutathione per­
oxidase and tocopherol in patients on continuous ambulatory
peritoneal dialysis (CAPD). Kidney int. 26:612 (1984).
Katirtzoglou, A.: Oreopoulos, D.G.: Husdan, H.; Leung, M.:
Ogilvie. R.: Dombros, N.: Reappraisal of protein losses in
patients undergoing continuous ambulatory peritoneal dialysis.
Nephron 26:230-233 (1980).
Rubin, J.; Nolph. K.D.; Arlania, D.; Prowant. B.; Fruto, L.;
Brown, P.: Moore, H.: Protein losses in continuous ambulatory
peritoneal dialysis. Nephron 2<Y: 218—
221 (1981).
Twardowski, Z.: Ksiazek, A.; Majdan.M.: Janicka, L.: Bochenska-Nowacka, E.; Sokolowska, G.; Gutka, A.; Zbikowska,
A.: Kinetics of continuous ambulatory peritoneal dialysis
(CAPD) with four exchanges per day. Clin. Nephrol. 75:119-130
(1981).
Rubin. J.; Nolph, K.D.: Arfania, D.: Brown. P.: Follow-up of
hourly peritoneal clearances with and without nitroprusside
augmentation in patients undergoing CAPD. Kidney int. 16:
619-633 (1979).
Kaysen. G.A.; Schonfeld, P.Y.: Albumin homeostasis in pa­
tients undergoing ambulatory peritoneal dialysis. Kidney int.
25:107-114(1984).
Williams, P.: Kay, R.; Harrison, J.; McNeil, K.; Pettit, J.; Kelman, B.; Mendez, M.; Klein, M.; Ogilvie, R.: Khanna, R.:
Carmichael, D.: Oreopoulos, D.: Nutritional and anthropomet­
ric assessment of patients on CAPD over one year: contrasting
changes in total body nitrogen and potassium. Perit. Dial. Bull.
7:82-87 (1981).
Young, G.A.: Young, J.B.; Young, S.M.: Hobson, S.M.: Hil­
dreth, B.; Brownjohn. A.M.: Parsons, F.: Nutrition, and delayed
hypersensitivity during CAPD in relation to peritonitis. Neph­
ron 43:177-186(1986).
Accepted: May 28,1986
Dr. G.A. Young,
Scientific Director,
Renal Research Unit,
General Infirmary,
Leeds LSI 3EX (UK)
Downloaded by:
Vanderbilt University Library
129.59.95.115 - 10/26/2017 11:20:14 PM
infections during the study, which could have increased
the clearance of proteins [2,16). The similarity of the rate
of protein clearance during overnight and daytime ex­
changes suggest that any effects of diurnal rhythms or
physical activity were small.
201