STUDIES OF THE EXCRETION OF POLYVINYLPYRROL- IDONE BY … · IDONE BY THE NORMAL HUMAN KIDNEY* BY H. CAMPBELL, P. 0. KANE, D. F. MUGGLETON AND INGRID G. OTTEWILL From the Research

J. clin. Path. (1954), 7, 252.

STUDIES OF THE EXCRETION OF POLYVINYLPYRROL-IDONE BY THE NORMAL HUMAN KIDNEY*

BY

H. CAMPBELL, P. 0. KANE, D. F. MUGGLETON AND INGRID G. OTTEWILLFrom the Research Laboratories, May & Baker Ltd., Dagenham, Essex

(RECEIVED FOR PUBLICATION OCTOBER 1, 1953)

In developing any plasma substitute the problemof finding a high molecular weight compoundwhich can temporarily take the place of the plasmaproteins is the major one, since the electrolyte com-position involves no fundamental difficulties. Thefunction of this high molecular weight compoundis to exert an osmotic pressure across the bloodcapillary walls and, hence, assist in controlling thefluid balance in patients in whom there has been aloss of plasma protein. Provided any given poly-mer is pharmacologically inert and shows notoxicity either immediate, due to allergic reactions,or long-term, due to permanent storage of theforeign body in the tissue, its value in a plasmasubstitute will depend largely upon the persistenceof the osmotic pressure which it exerts across thecapillary wall. This will depend upon the rate ofelimination of the polymer from the body and therate at which it passes through the capillary wallsinto the tissue fluid. The elimination mechanismswill, in general, involve excretion of unchangedpolymer by the kidney or by other means, and itsdegradation to smaller molecules. The passageof polymer molecules through both the kidney andthe capillarv wall will depend upon the shape,structure, charge, and molecular weight of themolecules present. The importance of molecularweight in selecting a polymer for use in a plasmasubstitute has recently been emphasized by Boyd,Fletcher, and Ratcliffe (1953), but not enoughattention appears to be given to the effect that theshape and structure of a molecule can have on itspassage through a membrane. Consequently, themolecular weight most suitable for use in a plasmasubstitute will vary with the type of polymer. Themolecular weight distribution of a polymer does notnecessarily remain constant after its infusion intothe body, since it is unlikely that all the elimination

* Presented in part at the XIIIth International Congress of Pure andApplied Chemistry, Stockholm, Sweden, 1953, and at the symposiumon plasma substitutes organized by the International Blood Trans-fusion Society, Geneva, Switzerland, September, 1953.

This is the second of a series of papers, the first of which waspublished in the Journal of Polymer Science, 1954, 12, 611.

mechanisms will be independent of molecularweight. For a complete understanding of the be-haviour in the body of any polymer, knowledgeof the variation of its molecular weight distributionwith time is important. In all published work onthe fate in the body of polymers, which are nor-mally foreign to the body, no attempts appear tohave been made to determine such changes inmolecular weight distribution. In the presentpaper some of the changes in molecular weightdistribution which occur in the case of the particu-lar synthetic polymer, polyvinylpyrrolidone(P.V.P.), used in these studies are described.

Ideally, after an infusion of a plasma substitute,the polymer would be eliminated from the body atsuch a rate that the osmotic pressure due to thepolymeric component decreased as the osmoticpressure due to the plasma proteins returned tonormal as the deficiency in these was made good.An exact balance would, in practice, be impossibleto achieve, but a rational approach to the problemof selecting the most suitable molecular weightrange of any particular polymer is based on aknowledge of the time taken for the restoration tonormal, after sudden depletion, of the plasmaproteins in human subjects. Unfortunately, thereappears to be insufficient data available at presenton the rate of regeneration of plasma proteinsLtnder conditions similar to those when an infusionof a plasma substitute might be given. Most infor-mation on the regeneration of plasma proteins inhuman subjects is for cases of malnutrition.Wallace and Sharpey-Schafer (1941) have shownin human subjects not suffering from shock thatregeneration of the plasma proteins was virtuallycomplete in three to 90 hours, but no similar studieshave been made on human subjects who are in aseverely shocked condition due to loss of blood.Ebert, Stead, Warren, and Watts (1942) have shownthat in dogs, severely shocked after considerableloss of blood which was replaced by an equalvolume of isotonic saline, regeneration of proteins

on October 2, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.7.3.252 on 1 August 1954. D

ownloaded from

http://jcp.bmj.com/

EXCRETION OF POLYVINYLPYRROLIDONE

was extremely slow. No reason was given for thisslow regeneration. It is not known whether thepresence of a synthetic colloid can affect theregeneration of the plasma proteins through itsproperty of maintaining the circulating volume.

In the absence of such information as wouldallow a theoretical approach to the developmentof a plasma substitute, it is of value to determinethe behaviour in the body of those plasma sub-stitutes which are clinically acceptable. Such aplasma substitute is " plasmosan," which was firstdescribed by Thrower and Campbell (1951). Theclinical value of this material has been shown byArden, Mandow, and Stoneham (1951), by Haler,Clift, Jameson, and Swinton (1953), and by Russell-Stoneham (1953). The present paper describesrenal excretion studies of the colloidal component,P.V.P.. from human subjects after infusion of" plasmosan." This polymer, originally examinedfor use in a plasma substitute by Hecht and Weese(1943), has since been further developed in Franceand in this country. Since much of the publishedwork on this type of plasma substitute relates toP.V.P. of inadequately defined molecular weightrange, conflicting results have been reported, andit must be emphasized that the results described inthis paper apply to only one grade of P.V.P. ofdefinite molecular weight distribution. A prelim-inary account of the present studies has alreadybeen given by Campbell (1953).

ExperimentalAt the hospital each patient, selected by the medical

staff as being free from any kidney disease, was infusedwith up to one bottle (540 ml.) of " plasmosan." Thetime of beginning and ending the infusion was recorded,together with the volume of "plasmosan" unused.One 24-hour sample of urine was collected up to the timeof the infusion. After the infusion individual samples of

urine were collected during an initial period of about twodays and then 24-hour samples for at least another ninedays. In some surgical cases dressings which mightcontain P.V.P. from the open wounds were retained. Inour laboratories all samples of urine were initiallyexamined by the tests given below and then the P.V.P.was extracted. The molecular weight distributions of theextracted samples were determined. The wound dressingswere also examined.

Details of Patients and Treatment.-Table I givesdetails of each patient together with notes on the loss ofP.V.P. by various means. Mr. G. P. Arden, in charge ofall the surgical cases, stated that there was no excessivebleeding in any case. The actual quantity of blood lostin the operation cannot, however, be stated.Plasmosan from the same batch was used in all

infusions. The composition of " plasmosan " haspreviously been given by Thrower and Campbell (1951).The molecular weight distribution curve for the P.V.P.used in this batch is given in Figs. 4 and 5, and had notchanged during storage since the preparation of the" plasmosan."

Experimental Methods.-The clarity, specific gravity(by hydrometer), and pH (by glass electrode) of eachsample of urine were noted. Albumin was tested forqualitativity by the sulphosalicylic acid test and forreducing sugar by Benedict's test. Creatinine wasdetermined by the Jaffe method, the resulting intensity ofcolour being measured on an EEL colorimeter using aNo. 624 filter. Details of the latter three tests have beengiven by Hawk, Oser, and Summerson (1947). TheP.V.P. concentration was determined colorimetricallyby the iodine method as described by Campbell andHunter (1953) and by Campbell, Kane, and Ottewill(1954a). An EEL colorimeter with No. 623 filter wasused to measure the intensity of colour. From the P.V.P.concentration, and the volume of the urine sample, theweight of P.V.P. present was obtained.A representative quantity of the P.V.P. in each sample

of urine was extracted by the following method. To theurine was added 22% by weight of exsiccated sodium

TABLE IDETAILS OF PATIENTS

Patient Sex Type of Kidney Function Known LossSerial No. Se Case Age Diagnosis or Operation from Clinical Of P.V.P.Evidence

3 M Medical 23 Suspected mildly active tuberculous pleural effusion Normal None13 F ,. 58 Duodenal ulcer ,.16 M 26 Pneumonia resolved before P.V.P. infused ,.17 M 9 31 Abdominal investigation-N.A.D. . D18 M ,, 61 Gastric ulcer ,, None4 F Surgical 73 Fractured left femur (Smith-Petersen pin) ,, B, C, E5 F ,, 54 Bilateral hailux valgus tKeller's) ., E6 M , 18 Removal of exostosis of first metatara1 B, E7 F ,, 46 Bilateral hallux valgus (Keller's) , E8 F ,, 53 ,, ,. ,, ., , C, E9 F ,, 50 ,, ,, ,, ,, , B. C, E10 F ,, 39 ,, , , , ,B,E,AI I F ., 33 , ,,,, B,E12 F ., 21 Bilateral KeUer's operation and spike arthrodesis C,C, E, A15 F ,, 47 Right hailux valgus (Keller's) , E

A-loss by menstruation, B=loss of urine by hospital, C=incontinence and/or bowels opening, D=loss of urine in laboratory, E=loss bybleeding into wound dressing.

S

253



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

H. CAMPBELL AND OTHERS

sulphate, which was dissolved by heating. The solutionwas cooled to 25-300 C. and the P.V.P. was thrown outof solution as an insoluble gel. The solution was extractedthree times with one-quarter its volume of chloroform.The separating funnel was shaken vigorously for fiveminutes for each extraction. Any gel formed was clearedon a centrifuge, and the "skin" formed at the interfacewas extracted with a little more chloroform to removeany adsorbed P.V.P. The chloroform extracts werecombined and, if opalescent, were cleared by the additionof a small quantity of methanol. This had the effect ofconverting the two-phase system chloroform-water intoa single phase, and sodium sulphate dissolved in theaqueous phase was precipitated. The clear liquid wasthen filtered and evaporated to dryness on a steam bath.The solid residue was freed from traces of chloroformby heating on a steam bath at reduced pressure (filterpump) and dissolved in an appropriate volume of waterto give an aqueous solution which was then suitable forthe determination of the molecular weight distribution.Experiments, in which P.V.P. of known molecular weightdistribution was added to urine, have shown that therewas no detectable difference between the dissolved andthe extracted P.V.P. Distributions of molecular weightwere determined by the turbidimetric method as describedby Campbell, Kane, and Ottewill (195 io).Each wound dressing was extracted with boiling water.

After filtering, the aqueous solutions were co-icentratedby vacuum distillation and clarified in a centrifuge. Thesolutions were then examined by the icdine method forthe presence of P.V.P. Attempts were made to determinemolecular weight distribution curves. A wound dressingfrom a patient who had not received " plasmosan " wassimilarly examined.Methods of Calculation. Since the actual period of the

infusion was at least one hour, zero time for calculatingthe rate of excretion of P.V.P. has been taken as the

arithmetic mean of the beginning and end of this pericd.This introduces an indeterminate error when a com-parison is made of the excretion by different patients attimes less than 10 hours after infusion, but this error isnegligible for longer periods.The uncorrected weight of P.V.P. present in any

sample of urine was obtained from the iodine estimationwithout allowance for variations in sensitivity of themethod with molecular weight. When a correction wasmade for this, using the peak molecular weight of thedistribution curve, the weight of P.V.P. present isreferred to as the corrected weight. The correctionfactors used are given by Campbell, Kane, and Ottewill(1954a). This may introduce a small error, since the truecorrection factor, which cannot at present be determined,will depend upon the distribution of molecular weightsrather than on the peak molecular weight.

In comparing the molecular weight distributicns of theinfused and total excreted P.V.P., it is convenient toconsider the fate in each patient of 1 g. of P.V.P. whichis completely characteristic of all the P.V.P. infused.Of each gramme infused, the fate of R'100 g. is knownwhere R is the percentage total weight excreted. There-fore, molecular weight distribution curves for 1 g. ofinfused and R'100 g. of total excreted P.V.P. are com-pared. This latter is obtained by summing the molecularweight distribution curves for r/100 g. of each successivesample of excreted P.V.P. (of known percentage weight r)where Er= R.

Experimental ResultsA complete picture of the renal excretion of

P.V.P. can be obtained only from those cases inwhich there was negligible loss of P.V.P. (seeTable I). The data from those patients in whichindeterminate losses of P.V.P. did occur can, how-

TABLE I1EXPERIMENTAL DATA FOR PATIENT NO. 16 AFTER INFUSION OF 16-5 G. P.V.P.

Determination of P.V.P. in Urine

Time from Sample PeakInfuson VlumeMolecularInfusonvlume Weight

(hours) (Ml.) X 103

1,090475435330220360540

1,8002,2502,034750

1,9351,6602,0402,7101,7701,3402.4202.6051,270

10-5202121-5252625-530 533-537

34363940 53947384959

Weight Uncorrectedweigto Weight ofFatrP.V.P. (g.)

1-481-081-081-071.05

1-041-041 001-001-00

1-001-001-001-001-00I-001-001-00I 00

0

4-894-35

0 990-200-190-211-140 930-51

0-33

0-230-140-130-110-080-020 030-020 02

CorrectedWeight ofP.V.P. (g.)

0

7-244-701-070-210-200-221-190-930-510-33

0-230-140-130-110-080-020-030-020 02

Laboratory Tests on Urine

Clarity

Normal

turbid

Normal

..

SpecificGravity15- C.

1-0131-0141-0151 -0241-0121-0071-0151-0071-0111-0121023

1-01210121-01510081-0081-017I -0091 0091-013

pH Albuminl

6-17-06-05-86-66-97-05-9 _6-5 _6-1 _5-8

6-16-4 _6-06-56-55-9 _636-462 _

254

SampleNumber

16 116 216 316 416/516 '616 716 816 916 1016,11

16, 1216 1316/ 1416 1516 1616 1716 1816/1916 20

Sugar

Pre-infusion0-5918-520-521-523-5456993117

141165189213237261285309333

Creatinine(g.)

1-310-220-640-710-100-120-151-391.911-771-77

1-821-681-771-711-501 771-791-801-46

_ ~ 1 I~~~ I_



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


ever, be used to support some of the main con-clusions. A set of data for one typical medicalpatient is given in Table II.

Pattern of Excretion.-A typical set of molecularweight distribution curves for P.V.P. recoveredfrom successive samples of urine passed by patient16 is given in Fig. 1. The molecular weight distri-bution curves for all the samples of recoveredP.V.P. showed the characteristics typical of frac-tionated material, a narrower molecular weightrange and sharper peak than with the originalunfractionated polymer. The curves, very sharpand narrow on the first day, gradually broadenedas the post-infusion period extended. The increasein peak molecular weight with time after the in-fusion is illustrated in Fig. 2. Towards the end ofthe excretion study on any particular patient, about150 hours after infusion, the difference between

successive molecular weight distribution curveswas less than the experimental error, which becamelarge in this period due to the extremely smallamount of P.V.P. present in the urine. The datafor the cases which were incomplete through lossof urine or other causes were not in disagreementwith this general picture.

Rate of Excretion.-Fig. 3 shows the variationof percentage weight excreted, with time after theinfusion, for four medical and three surgical cases.The excretion rate, high during the first 24 hoursafter infusion, decreased to a small value afterabout 100 hours. Excretion did continue at thissmall, but measurable, rate for 300 to 600 hours fordifferent patients. The rapid excretion immedi-ately after infusion is illustrated by Table III, whichincludes data for all patients. The variation be-tween different patients due to errors already dis-

25 50 75Molecular Weight (x I-3)

Fso. 1.-Molecular weight distribution curves for P.V.P. recovered from successive samples of urine from patient16 (see Table II). These curves are not normalized.

255



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


40302010

4030

, 20

X 10

0

0 403 30, 20' 100

30

::40

0m 302010

40302010.

0 100 200 300 400 500 600Time from Infusion (hours)

Fio. 2.-Increase in peak molecular weight with time after infusion. A is patient No. 9 (surgical), B is patient No. 12(surgical), C is patient No. 3 (medical), D is patient No. 16 (medical), E is patient No. 13 (medical).

TABLE IIIPERCENTAGE WEIGHT OF P.V.P. EXCRETED DURING

FIRST 48 HOURS AFIER INFUSION

% Weight % Weight %, WeightExcreted Excreted Excreted

Type of Patient at 12 Hours at 24 Hours at 48 HoursCase No.

Uncor- Cor- Uncor- Cor- Uncor- Cor-rected rected rected rected rected rected

- I_Medical 3

13

1617 (a)18

4462586658

56

7575

811

5570667365

668383

89

6475738173

758492

96

Surgical 4 (b)5 50 56 56 73 75 796 12 14 22 25 35 407 (c) I 1 13 17 48 528 12 15 22 27 32 389 17 19 20 22 23 2510 (b) -

I1 68 88 76 92 82 9812 32 40 39 47 44 5215 34 49 53 62 69 75

(a) Corrected weights are not available. (b) Data have not beencalculated due to considerable loss of urine in this period. (c) Thelow values at 12 and 24 hours are probably due to retention of urineover this period.

cussed on page 254, in the times calculated for eachsample of urine is small at 12, 24, and 48 hoursafter infusion.

An indication of the quantity of P.V.P. excretedover shorter post-infusion periods can be obtained(Table IV). Samples of urine, collected from twopatients (Nos. 13 and 16) within 15 minutes of theend of the infusion, contain the P.V.P. actually

TABLE IVEXCRETION OF P.V.P. DURING AND IMMEDIATELY

AFTER INFUSION

Infusion Timne of % Weight of P.V.P.Collection Recovered

Patien of UrineNo. Time Duration Samples Uncor Corrected

(hours) ~~rected

13 10 45 to 11-45 100 11-50 3-5 5-216 10-05 to 12-20 2-25 12 34 29-8 44-1

excreted by the kidney during the infusion and fora short period afterwards. These data also indicatean immediate and rapid excretion of P.V.P.

Total Recovery of P.V.P.-The total recoveriesof P.V.P. from all patients are given in Table V.Renal Threshold Molecular Weight.-In all the

patients studied, P.V.P. with molecular weights overthe whole range 0 to more than 100,000 was re-covered from the urine. If there is a threshold

A

- B-I + +

- +

+~~~~~

+E

I DI 1- I I .1 I I - I I I4- ~~~~~4.- -.+

II I I I I I~~~~~~~~~~~~

256



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


100 200 300 400Time from Infusion (hours)

500

FIG. 3.-The variation, with time, of percentage weight of P.V.P. recovered from urine.A + Patient No. 11 (surgical) B + Patient No. 18 (medical)0 Patient No. 5 (surgical) 0 Patient No. 16 (medical)x Patient No. 15 (surgical) x Patient No. 13 (medical)

& Patient No. 3 (medical)

TABLE V

TOTAL RFCOVERY OF P.V.P. FROM URINE OF PATIENTS

Known Loss of Y egtRcvrdfo rnType of Patient P.V.P. other % Weight Recovered from UrineCase Number than by Bleed-

ing from Wound Uncorrected Corrected

Medical 3 None 70-3 88-013 . 92 8 106-216 . 87-9 105-317 . 91-3 *18 . 86-0 107-4

Surgical 4 Some 13*8 14*15 None 81-5 89-86 Some 42-4 46-67 None 53-9 59-08 Some 38-7 44-39 . 27-1 29-210 . 48-2 57-111 . 90 5 107-312 ..567 64-915 None 76-1 84-4

Molecular weight distribution data not available to allow correc-tion to be made.

molecular weight above which molecules of P.V.P.cannot pass through a normal kidney, this thresholdmust, therefore, be considerably greater than100,000.

Degradation of P.V.P. in the Body.--When morethan about 80% of the infused P.V.P. was re-

covered a comparison could be made between themolecular weight distribution for the infusedP.V.P. and for the sum of all the samples of P.V.P.excreted by any one patient. No allowance hasbeen made for the departure of the recoveredP.V.P. from 100%'. Fig. 4 shows molecular weightdistribution curves for four medical patients andFig. 5 for two surgical patients. The distributioncurves are presented as histograms, since this sim-plified the calculations. Owing to the very smallquantity of P.V.P. of molecular weight above100,000, the histograms in this region cannot bedrawn so that they are distinguishable from thebase line.Wound Dressings.-The presence of P.V.P. in

the wound dressing was demonstrated by the iodinemethod of estimation and the turbidimetric methodfor the determination of molecular weight distri-bution, but no quantitative results were obtainedowing to the presence of interfering substances.

10

0 5VW

U

v

<, C

e

O+ ~~~~~~~~~~~A

O I I I I I I I- I

rg ~~~~~~~~~~~~~~~~~~~B

I ' I 1 I I I 1600

257

c



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


Molecular Weight (x10-3)FIG. 4.-A comparison of the molecular weight distribution curves of in

covered ( ...... ) P.V.P. for four medical cases. A is patient No. 3is patient No. 16, and D is patient No. 18.

100 0 25Molecular Weight (x 10-3)

FIG. 5.-A comparison of the molecular weight distribution curves orecovered ( ) P.V.P. for two surgical cases. A is patient

No. 15.

These were also found in dressings from patientsnot receiving " plasmosan."

DiscussionAlthough all patients had been selected by the

medical staff of the hospitals concerned as freefrom any kidney disease, some properties of the

urine samples, which are listedabove in the experimental sec-

B tion, were recorded as supple-mentary evidence. Apartfrom the 24-hour volume andcreatinine content, there ap-peared to be no marked de-partures from normality. Sur-gical cases differed from medi-c-al in that both urine volumeand creatinine excretion fluctu-ated considerably and, on thewhole, were appreciably lower

75 ioo than what are regarded as nor-mal values. Such variations.

D particularly in the immediatepost-operative period, mightwell be due to variations inthe patient's fluid intake anddisturbance of his nitrogenmetabolism. The administra-tion of P.V.P. itself has alreadybeen shown by Thrower andCampbell (1951) to exert no ill

l i effect on kidney function.

Low recoveries of P.V.P. in75 100 some of the patients studied

may have been due to one orifused( ) and re- more of the following: (1) Loss,B is patient No. 13,C of urine due to (a) inconti-

nence of urine or its contami-nation by faeces, or (b) failure

B to preserve the urine samplesor their loss during examina-tion. (2) Loss of blood intowound dressings and due tomenstruation. Such losseswere of most significance whenthey occurred in the first 48hours after infusion when theconcentrations of P.V.P. in

I0 ,blood and urine were greatest.75 100 The actual quantity of P.V.P.

lost cannot be determined. Inf infused ( ) and all the medical cases the con-o. 11 and B is patient stancy of the 24-hour volume

and creatinine content of theurine indicated that losses were negligible. In thesurgical cases, however, no such simple interpre-tation was possible, since the effect of surgery onthe urine volume and creatinine excretion was notknown. In all surgical cases loss of P.V.P. due tobleeding occurred. For reasons discussed above,such losses could not be assessed and, hence, the

258



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


anmount of P.V.P. reco\ ered must be regarded asminimal.

In all the cases studied, no differences betweenthe experimental results for male and femalepatients have been found. It is impossible to saywhether the experimental differences betweensurgical and medical cases, such as the total P.V.P.recovery (Table V), are due to any other causesthan the losses of P.V.P. known to occur in thesurgical cases.

Apart from its dependence upon blood concen-

tration, the rate at which P.V.P. passes through thekidney must depend inversely upon its molecularweight in order to explain the general shape of thecurves in Fig. 2 which give the increase in peakmolecular weight with time after infusion. Amore detailed analvsis of the kinetics of excretionis impossible at present owing to the complicatingfactor of degradation of P.V.P. Although P.V.P.passes through the kidney very slowly at highmolecular weights, evidence has been obtained thatP.V.P. of all the molecular weights present in thepolvmer infused can pass through the kidney. If a

threshold molecular weight exists, above whichP.V.P. molecules will not pass through the kidney,this is in excess of 100,000. The value of thisthreshold molecular weight cannot be stated atpresent with greater accuracy, since the actualweight of P.V.P. of molecular weight greater than100.000 in "plasmosan" is too small to allowprecise measurements to be made. An accuratedetermination would require excretion studies usingP.V.P. of a much higher molecular weight. Theexcretion of P.V.P. of all the molecular weightspresent in " plasmosan " is confirmed by the com-

plete recovery of P.V.P. in the urine of thosepatients in whom loss of P.V.P. may be regarded as

negligible (Table V). This complete recovery indi-cates that disposal of P.V.P. in the body other thanby renal excretion is insignificant and that it is notpermanently stored in the body. Histologicalevidence that the grade of P.V.P. used in " plas-mosan" is not permanently stored has been foundby Whitelaw (see Thrower and Campbell, 1951) inrabbits, and by Haler and co-workers (1953) in a

human subject. These experiments refer to plasmaconcentrations of P.V.P. not grossly in excess ofthose likely to be found in normal clinical practice.Conflicting evidence concerning storage of othergrades of P.V.P. in various animals has been re-

ported by Korth and Heinlein (1943), by Bargmann(1947), by Schoen (1949), by Fresen (1949a), andby Steele, Van Slvke, and Plazin (1952). This lackof agreement between different workers may partlybe due to the use of different grades of P.V.P., the

molecular weight ranges of which have never beenspecified, since Fresen (1949b) has shown that thepossibilitv of storage will certainly increase withincreasing molecular weight. Other relevant fac-tors may be the weight of P.V.P. per kilogram ofbody weight infused, and the possibility of differ-ences in the renal excretion between differentspecies of animals.The threshold molecular weight of excretion for

proteins as found by Bayliss, Kerridge, and Russell(1933) is about 68,000 and is much smaller than thevalue for P.V.P. The explanation of this may be,in part, due to differences in the shape and electriccharge of the molecules. Another difference does,however, exist between the plasma proteins andP.V.P. molecules in solution, which may be ofimportance. No protein molecule can uncoil, andhence change its cross-sectional area, withoutundergoing irreversible denaturation. Scholtan(1951) has shown that the P.V.P. molecule, likemost flexible, long-chain, synthetic polymer mole-cules, is, in solution, randomly kinked and, unlikethe protein molecule, continually undergoes revers-ible changes in the actual degree of kinking. Thereis a most probable configuration and cross-sectionalarea, which, for P.V.P. of a higher molecular weight,may prevent passage through the kidney. Revers-ible changes in the degree of kinking, however,could yield molecules of smaller cross-sectionalarea which might then be capable of passingthrough the kidney. It may be this mechanismwhich allows P.V.P. molecules of high molecularweight to pass slowlv through the kidney.A comparison of the histograms of Figs. 4 and 5

shows that there is more P.V.P. of a low and lessof a high molecular weight recovered in the urinethan was infused into the patient. Some form ofdegradation of P.V.P. molecules must, therefore,occur during their passage through the body, butthe mechanism of this change is at present un-known. These observations, which can be dueonly to a limited degradation of P.V.P., are not indisagreement with those of Steele and his colleagues(1952), who found that after infusions of C14 taggedP.V.P. onlv a very small amount of radioactivecarbon dioxide was expired.

Evidence is now available that P.V.P. is veryrapidly excreted in the 48 hours after an infusion.Table IV indicates that there is no delay in thebeginning of excretion. Table III shows thatwhere there is no loss of P.V.P. between 65 and90% (corrected weights) is excreted in the first 24hours after infusion. In the first 12 hours over50% is excreted. From the above figures it followsthat, 24 hours after an infusion, between 10 and

259



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


35% of the P.V.P. is left in the body. This mustrepresent the maximum present in the blood streamat 24 hours, since some of the P.V.P. may be, atleast temporarily, in other body fluids or tissues.

In conclusion, these excretion studies indicatethat the P.V.P. in " plasmosan" as at present con-stituted is not permanently stored in the body. Itwould appear that clinically beneficial results can

be obtained with a grade of P.V.P. which is excretedas rapidly as the present studies indicate. Whetheran improved plasma substitute would result fromusing a grade of P.V.P. which is excreted moreslowly can be decided only by further studies.

SummaryClinical studies are described of the renal

excretion of polyvinylpyrrolidone (P.V.P.) from15 patients.Over 90 0% of the infused P.V.P. can be accounted

for in the urine of those patients in whom there wasno loss of P.V.P. due to bleeding or to incompletecollection of urine.

If a threshold molecular weight exists abovewhich P.V.P. would not be excreted by the normalkidnev, this is in excess of 100,000.

There is a very rapid excretion of P.V.P. in thefirst 24 hours after an infusion.There is evidence for degradation of P.V.P.

during its passage through the body.

We are indebted to the following without w;iosecooperation in providiing us with the necessary specimens

these studies would have been impossible: Mr. G. P.Arden, Heatherwood Hospital, Ascot; Dr. F. Marsh,St. Margaret's Hospital, Epping; Dr. A. Piney, St.Mary's Hospital, Plaistow; Dr. H. W. Salmon, EastHam Memorial Hospital; Professor C. Wilson, TheLondon Hospital. We have been assisted in the measure-

ments in our laboratories by Mr. R. V. Crouch, Mr. D.Gilbert, Mr. S. W. Head, and Mr. T. Moore, and incomputing results by the calculating section of theAccounts Division of this firm.

This paper is published by permission of the directcrsof May & Baker Limited.

REFERENCES

Arden, G. P., Mandow, G. A., and Stoneham, F. J. R. (1951).Lancet, 1, 1099.

Bargmann, W. (1947). Virchow's Arch. path. Anat., 314, 162.Bayliss, L. E., Kerridge, P. M. T., and Russell, D. S. (1933). J.

Physiol., 77, 386.Boyd, A. M., Fletcher, F., and Ratcliffe, A. Hall (1953). Lancet, 1, 59.Campbell, H. (1953). Ibid.. 1, 148.

and Hunter, G. (1953). Ibid., 1, 197.Kane, P. 0.. and Ottewill, I. G. (1954a). To be published.

(1954b). J. Polymer Sci., 12, 611.Ebert, R. V., Stead, E. A., Warren, J. V., and Watts, W. E. (1942).

Amer. J. Physiol., 136. 299.Fresen, 0. (1949a). Zbl. Chir., 74. 65.

(1949b). Arztl. Forsch., 3, 308.Haler, D., Clift, A. F., Jameson, M. H.,rand Swinton, C. F. (1953).

Med. Press, 230, 131.Hawk, P. B., Oser, B. L., and Summerson, W. H. (1947). Practical

Physiological Chemistry, 12th ed., Blakiston Co., Philadelphia.Hecht, G., and Weese, H. (1943). Munch. med. Wschr., 90. 1 1.

Korth, J., and Heinlein, H. (1943). Arch. klin. Chir., 205, 230.Russell-Stoneham, F. J. (1953). Plasma Substitutes Symposium,

organized by the International Blood Transfusion Society,Switzerland.

Schoen, H. (1949). Klin. Wschr., 27, 463.Scholtan, W. (1951). Makromol. Chem., 7, 209.Steele, R., Van Slyke, D. D., and Plazin, J. (1952). Ann. N. Y. Acad.

Sci., 55, 479.rhrower, W. R., and Campbell, H. (1951). Lancet. 1, 1096.Wallace. J., and Sharpev-Schafer, E. P. (1941). Ibid., 2. 393.

260



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Documents

STUDIES OF THE EXCRETION OF POLYVINYLPYRROL- IDONE BY … · IDONE BY THE NORMAL HUMAN KIDNEY* BY H. CAMPBELL, P. 0. KANE, D. F. MUGGLETON AND INGRID G. OTTEWILL From the Research