Chemical measurements of inulin concentrations in peritoneal dialysis solution

103

Clinica Chimica Acta, 76 (1977) 103-112 @ Elsevier/North-Holland Biomedical Press

CCA 8418

CHEMICAL MEASUREMENTS OF INULIN CONCENTRATIONS PERITONEAL DIALYSIS SOLUTION

PAUL

BROWN

and KARL

D. NOLPH

IN

*

Department of Medicine, Division of Nephrology, University of Missouri Medical Center and Veterans Administration Hospital, Columbia, MO. 65201 (U.S.A.) (Received

November

5th, 1976)

Summary Peritoneal inulin clearance during peritoneal dialysis is a useful indicator of efficiency of removal of larger solutes. Peritoneal dialysis solutions usually contain high glucose concentrations that interfere with the chemical measurements of inulin concentration. In these studies, inulin recoveries in simple glucose and peritoneal dialysis solutions with three chemical methods were compared; all methods reportedly were subject to minimal glucose interference. The findings showed one method subject to minimal and predictable glucose interference in all solutions tested. For two methods, interference in peritoneal dialysis solutions exceeded that predicted from glucose alone; this unexpected interference was shown to most likely result from carmalized glucose subsequent to the routine autoclaving of peritoneal dialysis solutions.

Introduction It has been proposed that retained solutes in the molecular weight range of 500-5000 may explain some manifestations of the uremic syndrome [l]. In clinical studies of dialysis efficiency, investigators have attempted to include measurements of the removal rate of solutes in this “middle molecular weight” range. Inulin (5000 daltons) is a solute in this molecular weight range, and is available for intravenous administration. In subjects with impaired renal function, serum concentrations of inulin decrease very slowly over several days following a single injection. It is thus suitable for clearance studies during dialysis. Also, whereas clearances of very small solutes such as urea and creatinine are marked* To

whom

correspondence

should

be addressed.

ly blood flow and dialysate flow dependent, inulin clearance is primarily limited by t,he area and permeability (*harac%cristics of the membrane [2]. Many investigators are now interested in possible factors that may alter the effective area and/or permeability of the peritoneal membrane; inulin clearances provide an important tool for such investigations. Although mulin clearances might. be done most accurately wit,h radiolabeled inulin, approval to administ,er radioactive materials intravenously in clinical studies is not readily obtained. In contrast, nonradioactive inulin is readily available for int,ravenous injection; however, many chemical measurements of inulin in peritoneal dialysis solutions are subject to glucose interference. Dialysis solutions usually contain glucose in large concentrations (0.5 to 4.25 g/dl). The basis for most inulin determinations is the hydrolysis of the inulin molecule with a st,rong acid followed by quantitative analysis of the hydrolysis products (mainly fructose). Fruct.ose levels are deWmined either by their reducing power or by t,heir reaction with a specific color-forming reagent. Several of the color-forming reagents used have I)een ,‘I-indo!e acet,ic acid [ 31, anthrone [ 41. resorcinol [ 5,6], cysteine hydrochloride [ 7 1, and diphenylamine [g-16]. We have recently evaluatetl several of these methods in our laboratory to compare their applicabilities for inulin determinat.ions in peritoneal dialysis solution. Each of the methods tested claimed a high specificity for fructose and hence a minimum glucose int,erferencr. Our studies of these met,hods showed the interference in aqueous glucose solutions to be low; but there was a marked increase in interference when dialysis solutions with the same glucose concentrations were subst,ituted for glucose solutions in water. Studies were undertaken to determine the source of interference and to ascertain which method resulted in the least total interference. Methods

Evaluation

of methods

claiming low glucose interference

Studies included the evaluation of three methods. The first, the rnethod of Heyrovsky [3], involves the use of HCl to hydrolyze inulin. 8-Indoleacetic acid is used as the color forming reagent. In the presence of hydrochloric acid it combines with fructose to form a deep purple color when allowed to set for several hours at room temperature. The color formed is highly stable and interference from glucose is reported to be on the order of 0.5-1.0%. The second evaluation was that of Messineo and Mussarra [ 71. They describe two methods. The first uses cysteine hydrochloride in sulfuric acid to form a green chromophore with a fructose derivative. The second method, the one tested, is more sensitive and involves complexing of the green chromophore (formed with the cysteine-hydrochloride sulfuric acid reagent) with tryptophan to form another chromaphore which is pink. According to the authors, aldohexoses do not interfere with the fructose determination because under the conditions stated they do not react even when levels as high as 500 mg/dl are present. The third method evaluated was the method of Walser et al. [ 161. Hereafter called the Walser method, it was derived from a group of previously published methods [ 10,17,11,15]. This method measures the alkali-stable portion of the

inulin molecules described by Cotlove [17,18]. The alkali stable portion is a comparatively larger and more homogenous molecular weight fraction of the polymers within the total inulin population. Evaluation of these methods included the specific recommended procedures and spectrophotometric measurements using solutions of glucose in water over a concentration range of 100-1000 mg/dl. The sprctrophotometric at)sorptions with peritoneal dialysis solution using each procedure (but without addition of inulin) were also determined. Because of the observed interference characteristics, evaluation of the Heyrovsky method also included studicas in test solutions containing individual componc>nts of peritoneal dialysis solution. The test solutions studied included sodium acetate, 610 mg/dl; sodium chloride, 560 mg/dl; calcium chloride, 29 mg/dl and sodium hisulfite, 10 mg/di, respectively. One other test solution contained all of the above concentrations in water without glucose, thereby, in essence, being peritoneal dialysis solution with no glucose added. Simultaneously with spectrophotometric absorption studies on all of the above, following application of each chemical procedure, the absorbance of solutions containing inulin and water at various concentrations were determined. Because peritoneal dialysis solutions are heated or autoclaved in preparation, studies of these methods were also performed ‘&IIglucose solutions before as well as after varying intervals of heating at 100” C.

Further studies of the Waker method Because the Walser method emerged as the method least subject to interference, additional studies of t,his assay were performed. ‘These studies included increasing and decreasing the recommended heating time with alkali, increasing and decreasing the heating time after the addition of diphenylamine reagent, and a study of the effects of rapid cooling in an ice bath compared to gradual cooling at room temperature. Inulin recoveries using a dialysate blank and inulin standards in water to determine known amounts of inulin in dialysate were also examined using the modified method as in the appendix. Recoveries were also determined from serum. Results

Comparisons

of the three methods

Fig. 1 shows comparisons of absorbance of different glucose concentrations for all three methods with glucose in water and from glucose concentrations made up from various dilutions of dialysis solution. Using the Heyrovsky method, it was found that aqueous glucose solutions did demonstrate absorption. The relationship was essentially linear and, at concentrations of 600 mg/dl, glucose gave the same absorption as 2 mg/dl of inulin in water. Actual dialysate samples gave absorption readings much higher at given glucose concentrations than predicted from the results with glucose with water. Studies of the Messineo and Mussarra method gave similar results. Glucose in water solutions gave absorbance readings comparable in range and magnitude to those of the previous method. Dialysate again gave excessive absorbance above and beyond that predicted by the glucose concentrat,ion.

106 2.5

1

Messineo and Mussarra 11977)

Absorbance 518 nm 1.0

_D_-e-

_i)--0

0.5

Antonin Heyrovsky119561

1.5 Absorbance 530 "Ill 1.0 __C---

__o-

0.5

2.5

Walser.Davidsonand Orlolf 119551

2.0 I

Glucoselmgldll

Fig. 1 relates absorbance (vertical axis) to glucose concentration (horizontal axis) for the three methods as indicated. Studies with glucose in water are representrd by open circles connected by dashed lines and studies of various dilutions of peritonral dialysis solution are shown in solid circles connected by solid lines.

The Walser method also demonstrated some, but relatively much less, absorbance by glucose and water solutions; in contrast to the other methods, the absorbance of dialysis solution was essentially that predicted from the studies in glucose and water.

Further studies of the Heyrovshy

method

All of the test solutions containing the individual components of peritoneal dialysis solution including the mixed sample, free of glucose, gave readings essentially the same as that of a distilled water blank. A stock solution of 1000 mg/dl glucose gave an absorbance corresponding to the absorbance of a solution of 5.6 mg/dl of insulin in water. Dialysis solution diluted to a glucose concentration of 750 mg/dl yielded an absorbance corresponding to an inulin concentration of 10 mg/dl in water. Heyrovsky noted that heating the solutions in an attempt to speed the reaction caused an increase in glucose interference. Fig. 2 shows the results of heating a glucose solution containing 100 mg/dl in water. It can be seen that increasing the heating time increases the absorbance even if the solution is heated before the addition of the acid and color reagent.

Further studies of the Waker method With

the

Walser

method,

it was also found

that

increasing

or decreasing

the

107 1.0 0.9. 0.8 O.l0.6Absorbance 530 nm

o.5 0.4 0.3 0.2 0. 1

I 10

5

Fig.

2 shows

(vertical

studies

axis)

with

is rrlated

to

20

15 Minutes

75

$0

of Heatina

glucose

and

minutes

of

watm

(100

heating

mg

per

(horizontal

dl),

using

axis)

the

at 1OO’C

Heyrovsky prior

to

method.

the

Absorbance

determination.

heating time in alkali, seemed to have little effect on absorbance of a 5 mg/dl stock solution of inulin. Absorbance of undiluted dialysate (1500 mg/dl glucose), however, did vary with the heating time with alkali. These findings are depicted in Fig. 3. The optimum heating time following the addition of the diphenylamine reagent (Walser method) was examined and found to be 25 min. Fig. 4 shows the absorbance of a 5 mg/dl inulin stock solution with increasing heating time after the addition of the color forming reagent. Although maximum absorbance is not reached at 25 min, glucose interference continues to increase following the 25-min period. All further determinations with the Walser method including recovery ex-

.55 .50

-

45 -

Absorbance 620 nm

.30 .75

1 I

5

10

15

20

25

Minutes Fig.

3.

solution linr

Absorbancc to

rcpresrnt

studies

with

br

tested studies

a solution

(vertical

axis)

following of

dialvsate

of inulin

I

I

30

35

of Heating

the

using

with

I

8

8

&I

45

50

the

addition (1500

in water

Walser of

mg (5

I

55

60

NaOH

mg

method.

NaOH

per per

dl), dl).

as in the and

solid

is related assay. points

to Open

the

minutes

circles

conncctcd

of

connected by

solid

heating by lines

of

the

a dashed represent

perimcnts and those reprrsentcd in Fig. 1 involve a 15-min heating period in alkali and other steps as in the appendix. Fig. 5 shows the results of the stud& on the effects of rapid cooling. Two sets of inulin standards were run according to the Walser method, and at the end of each of the two heating periods, one set was immersed in an ice water

109

35 lllul1n Recovered imgldli

3o 75

lnulln Addedimgidll

lnulin AddedImgldl) Fig. 6. lnulin (horizontal).

recovery

Fig. 7. Inulin L.orlM).

reccwery

in dialysate

in serum

(vertical

(vertical)

using

using

Waker

Waker

method

method

is related

is related

to inulin

to inulin

added

added

to dialysate

to serum

(hori-

bath and the other was allowed to cool slowly at room temperature. Absorbances were calculated from these standards. It is almost impossible to distinguish the open from the closed circles and at the higher inulin levels, i.e. there is only a slight divergence. Figs. 6 and 7 show the results of inulin recovery experiments in dialysate and serum using the Walser method. Dialysate blanks were used to correct for glucose interference. Serum blanks were also used to correct for small amounts of glucose or other interferring factors in serum. The results show good recovery from both serum and dialysate. Mean recovery in dialysate and serum were 103 t 2 (S.E.M)% and 103 * 1 (S.E.M.)% respectively. With the other methods, the high interference of dialysis solutions made inulin recoveries very inaccurate. Discussion The results demonstrate some interference from glucose for all methods tested. The interference in dialysis solutions corresponds to that predicted from glucose only with the Walser method. The studies involving the heating of glucose solutions strongly suggest that caramelization of glucose may explain the interference phenomenon in the other methods explored. Peritoneal dialysis solutions are routinely autoclaved prior to marketing. Although the Walser method is subject to interference, the interference is relatively low and is linearly related to the glucose concentration. Using dialysate blanks, inulin recoveries in dialysis solutions are acceptable. During peritoneal dialysis, the amount of glucose absorption from dialysis solutions is quite similar in each patient from exchange to exchange [ 191. With

110 TABLE

I

CLINICAL

ASSESSMENT

Patient

Number exchanges

of

OF

“BLANK” Mean

glucose

in drainage

VARIABILITY concentration (mg/dl)

Standard

Standard

deviation

error

1

29

795

52

9

1

36

676

24

4

1

38

808

25

4

2

30

830

41

7

therefore the glucose concentration of any starting glucose concentration, dialysate drainage is quite similar. We have measured glucose concentrations in dialysate drainage from 133 exchanges during 4 dialyses in 2 patients. Mean values for each dialysis (20-38 exchanges per dialysis, each exchange with a 1.5% dextrose dialysis solution, 2 1 volume, 10 min inflow, 30 min dwell, 30 min drainage) are shown in Table I. The maximum standard deviation was 52 mg/dl during any dialysis corresponding to a 0.01 nm difference in absorbance or a 0.2 mg/dl change in the inulin blank correction. This would represent only 6% of the usual dialysate drainage absorbance or “blank” inulin concentration. A dialysate blank from drainage prior to administration of inulin should suffice as a correction blank for those inulin-containing exchanges which follow provided the glucose concentration in the pre-instillation solution remains fixed. Another more precise approach would be to measure glucose concentrations simultaneously in all dialysate samples and correct for the interference from a standard curve. To date, we feel the method of Walser et al. with appropriate corrections for glucose interference offers the best technique for measurements of inulin concentrations in dialysis solution and thus for determinations of peritoneal inulin clearances in a clinical setting. A detailed description of the method of Walser et al. with the modifications we suggest for peritoneal clearance studies is shown in the appendix. Appendix Suggested modifications

method for determination of inulin from the method of Walser et al.

in dialysate

and serum.

Minor

Required reagents 4 M NaOH. 0.75 M NaOH. ZnS04 reagent. 100 g ZnS04 . 7 H,O and 40 ml 3.125 M H,SO, made up to 1000 ml with distilled water. Diphenylamine reagent. 14 g diphenylamine (Baker) dissolved into 600 ml glacial acetic acid, 360 ml cont. hydrochloric acid is added to this mixture. The HCl should be added slowly with constant mixing in a fume hood. Preparation Dialysate

of samples samples are

assayed

with

no

preparation.

A dialysate

drainage

111

blank should be obtained before inulin administration. Serum is deproteinized according to Somogyi [20] prior to assaying. 7 ml I-I,0 (distilled) are added to 1 ml of serum. 1 ml &SO, reagent is added to the diluted serum. 1 ml 0.75 M NaOH is added dropwise during which time the solution is continually shaken. The tubes are allowed to set for 20 min and shaken occasionally during this period. After 20 min the tubes are centrifuged and the clear supernatant drawn off for analysis.

Determination

of inulin

Duplicate 2-ml aliquots of dialysate sample, deproteinized serum, inulin standards and distilled water are placed in clean clearly marked test tubes (20 X 150 mm). 0.5 ml 4 M NaOH is added to each tube which is then shaken, covered with a marble and placed in a suitable rack. The rack is lowered into a boiling water bath at 100°C and allowed to heat for 15 min. After the 15-min heating period the rack is removed and the tubes allowed to cool to room temperature. Upon cooling 6.25 ml diphenylamine reagent is added to each tube, the marbles are replaced and the tubes are again heated to 100°C for a period of 25 min. The rack is removed and the tubes again allowed to cool to room temperature after the 25-min heating period. When cool the absorbance is read in a spectrophotometer at 620 nm. The instrument is calibrated to 0 using the distilled water blank. The inulin concentrations in serum and dialysate are determined from the standard curve prepared from absorbances of inulin standards run simultaneously with the samples. The absorbance of serum and dialysate blanks (pre-inulin administration) should be substracted from the absorption of serum and dialysate samples respectively.

Calculations Inulin clearance is calculated by multiplying dialysate inulin concentration divided by serum inulin concentration with drainage volume divided by time the exchange. This calculation gives the clearance in ml/min.

of

Acknowledgement This 52216.

study

was supported

in part

by U.S.

Public

Health

Service

No.

1 AM-

References Babb.

A.

Organs.

Popovich,

K.P.,

Henderson.

L.W.

and

Hr)brovsky,

A.

(1956)

White. Roe,

Christrophrr.

T.G.

and

Scribncr.

B.H.

(1971)

Trans.

17,81&91

K.P. J.H..

and Epstein,

Nolph.

Samson. .J.H.

Clin. F.E. apd

K.D.

(1969)

Chin]. (1954) Goldstein,

Acta J.

.J. Clin. 1, Lab. N.P.

Invest.

48.

992-1001

470--474 Clin. (1949)

Med. J.

43, Biol.

475-478 Chem.

178,

839-845

Am.

Sot.

Artif.

Intern.

112

6

Kruhoffer.

7

Messineo,

L. and

8

Bojenscn.

E.

9

Rolf,

10

P. (1946)

D.,

Little,

(1952)

Surtshin,

J.M.

(1949)

11

Harrison,

H.K.

12

Corcoran,

A.C.

13

Alving,

14

Herbert,

15

Kennedy,

16

Waker,

77

Cotlovr,

18

Cotlovc,

19

Nolph.

20

Somowi,

AS., T.J.

Physiol. h1t.d.

Sand. H.1,.

J.

Biol.

Chcm.

180,

Proc. Page, J.

Yor. I.H.

and

Biochpm. Kleh,

Davidson,

J.

.I. (1953)

D.G.

and

Am.

.J. Phvsiul.

E.

(1954)

Fed.

Proc.

K.D..

Rosenfeld.

M. (1930)

P.S.. .J. Biol.

Biol.

B.1,‘.

13,

815.

Sot. 49,

Chvm. J.

282

Exp.

RIOI.

hlrd.

73,

351.-354

111 127.

~114 GO1 A$08

kol.

Chcm.

32,

90-m95

127.

6O!J

Gl G

819 In\rat.

.I. (1956)

171;.

39G

.J. Clin.

Invest.

34.

1520

1623

410

30

I’ov.ell, Chum.

Med.

(1939)

.J. Clin.

(1954)

Proc.

691499 275-

747~-754

Orloff.

E.

3, 266,

SuppI.

.I. Biol.

32,

5

Biochvm.

(1949)

Exp.

(1939)

.Millcr.

J.

142.

White,

1l-1

11

lnt.

and

(1938) and

&and.

(1972)

A.

(1942) and

E.

Acta

Rubin.

E‘.K. M.,

Acta

Mussarra.

J.T. 86.

and

G55+63

Danforth,

E.

(1970)

Am.

.J. Med.

Sci.

259,

272-281