Arch. Dis. Childh., 1966, 41, 229.

Paediatric Nephrology: Scientific Study of Kidneysand their Diseases in Infants and Children*

HENRY L. BARNETTFrom the Department of Pediatrics, Albert Einstein College of Medicine,

and the Bronx Municipal Hospital Center, New York, U.S.A.

The first George Frederic Still Memorial Lecturewas delivered in 1943 by S. P. Bedson, Professorof Bacteriology at the London Hospital. Inexpressing his appreciation for being asked toinaugurate this lectureship, he described his particu-lar satisfaction that the choice should have fallen ona laboratory worker since it carried with it theassurance that research of the kind he had beendoing, despite its academic flavour, was not withoutappeal to those engaged in clinical medicine. Itmay be a measure of one of the changes that hasoccurred in academic medicine in the intervening22 years that with equal fervour I must express mysatisfaction that the choice for the lecture this yearhas fallen on a paediatrician and clinical investigator,who is reassured that research of the type he hasbeen doing, despite its clinical flavour, is still notwithout appeal. My aim, in fact, in this lecturewill be to attempt to demonstrate how a primary andcontinuing interest in paediatrics served as thestimulus for a series of investigations of develop-mental renal physiology and of renal disease ininfants and children.

First of all, however, I should like to express myown deep appreciation to the British PaediatricAssociation for having asked me to give the 12thGeorge Frederic Still Memorial Lecture. In oneof the series of Pediatric Profiles in the Journal ofPediatrics, Sir Wilfred Sheldon (1956) describedStill's major role in the establishment of paediatricsas a separate field of inquiry and practice in thiscountry. Perhaps some of the differences betweenboth academic paediatrics and paediatric practice inour two countries can be explained by the fact that acomparable pioneering effort had been made some3; years earlier in the United States by Abraham

* The George Frederic Still Memorial Lecture, delivered at theAnnual Meeting of the British Paediatric Association in Scarborough,April 22, ) 965.

Jacobit. How comforting to realize that in onerespect our history is older than yours. Still'spublished work, like Jacobi's, dealt for the most partwith clinical, curative medicine, representing, aspointed out by Professor Crewe (1955), interestsentirely in accord with the stage of development ofpaediatrics and with the needs of that time. Thissolid, clinical foundation was necessary for andpermitted the emergence of the next major phase ofpaediatrics in which physiological and biochemicalexplanations were sought in an attempt to understandthe mechanisms of development and of disease.The early contributors to this subsequent phase ofpaediatrics could be considered successors to Still,and I should like to take this opportunity to ack-nowledge the outstanding contributions of one ofthem, Professor Alexis F. Hartmann, Sr., who diedlast September. Professor Hartmann was a corres-ponding member of this association. Your Presi-dent, Professor Gaisford, and I were both studentsof his, and with many others throughout the worldremember him with warm affection and deeprespect.My subject today concerns infants, children, and

their kidneys.There may be some who will challenge my

espousing the cause of paediatric nephrology bymaking it the subject of this lecture. If so, I mustadmit that the decision to do so was made with somereluctance. Our ambivalence about the use of thisand similar paediatrically diversive terms reflectsour uneasiness about the developments of more andmore specialization within paediatrics. On the onehand we seek to preserve the concept of the unity ofthe child and to understand his interrelationshipswith his family and with society. On the other, the

t In 1906 Still became the first Professor of the Diseases ofChildren in the United Kingdom at King's College. Jacobi hadbeen appointed Clinical Professor of Diseases of Children at theCollege of Physicians and Surgeons of Columbia University in 1870.

229

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Henry L. Barnettcomplexities of paediatric cardiology, neurology,haematology-and nephrology, among others-havebecome so great that productive research, mechan-ized teaching, and optimal clinical care of thespeciality's complicated patients all require ever-increasing specialization. Acceptance of paediatricnephrology acknowledges the scientific necessity forsuch specialization and, indeed, welcomes theadvances which have made it necessary. Its effecton general paediatrics is a very important butseparate matter.My purpose here then, in a sense, will be to

present the case for paediatric nephrology, not as anisolated speciality, but one that, with others, mustbring its specialized knowledge to general paediatricsand bridge it with both foetal and adult medicine.My evidence will consist ofinvestigations of develop-mental renal physiology and its relevance, and ofobservations on infants and children with renaldisease. Examples will be drawn for the most partfrom work done together with Dr. Chester M.Edelmann, Jr., and our associates in the paediatricrenal and electrolyte research unit of the AlbertEinstein College of Medicine and in the paediatricrenal clinics ofthe Bronx Municipal Hospital Center.

Developmental Renal Physiology and itsRelevance

A large share of our knowledge of developmentalrenal physiology rests upon the work of a member ofthis association, Professor R. A. McCance. Fromhis work with Widdowson and their associates, fromsome of our own work, and from that of manyothers, it is now possible to describe in some detailthe functional status of the kidney in newborninfants and its postnatal development. It has beenknown since 1940 (Bamett) that glomerular filtra-tion rate in young infants is low relative to a numberof structures and functions of the body includingsurface area, volume of extracellular fluid, and ratesof consumption of oxygen, calories, and fluid. Ithas been shown also that the relation betweenglomerular filtration rate and many tubular functionschange as the infant matures. Although a great

deal has been written about these changing relations,referred to loosely as renal immaturity, it is doubtfulwhether their true developmental significance isunderstood. However, something is known abouttheir consequences, many of which are of majorclinical importance.

Drug dosage and drug poisoning in infants.Enthusiasm for the enormous benefits gained fromthe discovery of many new classes of effective drugshas been tempered by an increasing concern aboutdrug reactions, many of which have been unexpectedand some desperately serious. Increasing knowl-edge of their mechanism of action promises thedevelopment of an even greater number of drugs inthe future with an accompanying increase inadverse reactions. As paediatricians we mustcontinuously be concerned with and responsible forassessing the sometimes narrow balance between thebenefits and dangers of drugs used in infants andchildren and be aware also of the potential danger tothe foetus of drugs given to pregnant women. Animportant aspect of this subject of developmentalpharmacology concerns renal handling of drugs inyoung infants. As an example (Table), the clear-ance of penicillin G in premature infants (Barnett,McNamara, Shultz, and Tompsett, 1949) is about90 ml./l *73 sq. m. compared with a value of 560 ml.in children. If rates of absorption and metabolismand the volume of distribution were the same inthe two groups, a dose of penicillin given on thebasis of surface area should, at equilibrium, produceconcentrations in blood about 6 times as high ininfants as in children. The larger volume ofdistribution associated with the proportionatelylarger volume of extracellular fluid in younginfants would lower this value by a factor of 1 * 5 to 2.A slower rate of absorption would lower the concen-tration further but would maintain it longer. Nowit happens that in most circumstances high andprolonged concentrations of penicillin in blood aredesirable. With drugs of greater toxicity, however,unexpectedly high concentrations in body fluids maybe dangerous. In anticipating from a given dose

TABLERenal Clearances of Inulin and Penicillin G in Premature Infants and Children

Mean Clearances (ml./mnin./l -73 sq. in.)No. Age Weight (kg.) Height (cm.) Inuln (ml./m i G_I | ~~~~~~ ~ ~~Inulin Penicillin G

Children12 I 3-12 yr. 13-46 95-150 127 560

.I (93-145) (311-954)Premature Infants

4 4-8 dy. 2*2-2*3 47-47*5 44 90(31-53) (72-114)

23()

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Paediatric Nephrologywhat the concentration in body fluids of a drug willbe, it is important to consider the developmentalfact that the renal handling of the drug may bedifferent in young infants; and that in most if not allinstances the rate ofrenal excretion will be decreased.

Altered renal function in the young infant mayassume even greater importance in accidental drugpoisoning, another increasingly serious paediatricproblem. A tragic example of this was the hospitaldisaster involving mass accidental salt poisoning ininfancy, reported in 1963 by Finberg, Kiley, andLuttrell. Owing to an error, sodium chloride wasused in place of cane sugar in the preparation offormulae which subsequently were estimated tocontain approximately 1,000 mEq sodium/litre. 14infants, 12 ofwhom were newborns and 2 less than 3months, received one or more of these salt feedings:6 of the 14 died, the first 5 before the nature of theproblem was realized. The highest concentrationsof sodium in serum ranged from 162 to 274 mEq/l.in the 8 infants in whom it was measured. 4 of theinfants were treated by peritoneal dialysis; in 2 ofthese it was believed that survival would have beenunlikely without it.

For several obvious reasons it is not possible tocompare these experiences with what might havehappened to older children had they receivedequivalent excesses of salt; it is not even possible tosay what would constitute equivalent doses. Bothfrom clinical experience and from what we do knowabout developmental renal physiology, however, wecan be fairly certain that the results would have beendifferent and far less disastrous. Developmentaldifferences in other physiological functions, especi-ally of the cardiopulmonary and neurocirculatorysystems, may make newborn infants more suscept-ible than older children to the harmful effects of thistype of salt poisoning. Nevertheless, these observa-tions do serve to demonstrate that the stage ofdevelopment of renal function in young infants, soadequate under normal circumstances, may imposea serious handicap under the types of stressesimposed by drug poisoning. It would not conflictwith developmental concepts if it turned out thatphysiological functions, harmoniously organized tohelp the young infant to grow and develop, might,under abnormal environmental conditions, createimbalances and impose handicaps.

Fluid and electrolyte therapy and infantfeeding. Observations on the development ofrenal function in young infants, especially concentra-ting mechanisms, have provided an important part ofthe physiological basis for clinical practices involvingboth fluid and electrolyte therapy and certain aspects

1400-

1200-

L 8o000._

, 8000cL

E 600-

° 400-

200-

0-

TOTALSOLUTES

ADULT

TOTALSOLUTES

INFANT

FIG. 1.-Renal concentrating capacity in infants andadults. Although measurements indicate that the meanvalue for maximum total osmolarity in adults is actuallybetween 1,100 and 1,200 mOsm./l. and that the low valuefor infants is due to a low rate of excretion of urea, thefamiliar (though inaccurate) values of 1,400 and 700 areused here for the sake of the argument being developed.

of infant feeding. It was shown several years ago(Fig. 1) that when infants were deprived of water fora period of 12 to 14 or 16 hours, the concentration ofosmotically active solutes in the urine increased to amaximum of about 700 mOsm./l. in striking contrastto values as high as 1,400 achieved by adults undersimilar circumstances. This relatively simple phys-iological observation, suggesting that young infantshave difficulty concentrating solutes in the urine,was applied to a variety of clinical situations.Many of these applications, as well as the interpre-tation of the original observation, are now known tobe incorrect.

Physiologically, the observation was interpretedas indicating either decreased elaboration or releaseof antidiuretic hormone, or decreased responsivenessof the renal tubules of the young infant. Itsrelevance to parenteral fluid therapy was based onthe calculation that at a maximum urinary osmolalityof 1,400 mOsm./l., the minimum volume of water inwhich 1 mOsm. of any urinary solute could beexcreted would be 0 - 7 ml. At a maximum concen-tration of only 700 mOsm./l. this value would bedoubled, 1-4 ml. of water being required for theexcretion of 1 mOsm. solute. It was argued, there-fore, that under clinical circumstances where solutesneeded to be excreted but water conserved, as in theinfants with salt poisoning, the young infant,requiring more urinary water to excrete excesssolute, would be at a disadvantage. Taken to-gether with other aspects of water and solutephysiology, these considerations suggested theadvisability of using hypotonic solutions containing100 to 200 mOsm./l. instead of isotonic saline in

231

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Henry L. Barnett

1400O

1200-

1000iL.4.3

EXPERIMENTALHIGH

PROTEINDIET

800. USUAL

DIETS

aI 600-

E 400. NON- NON- NON-

SOLUTES SOLUTES SOLUTES

ADULT INFANT

FIG. 2.-Renal concentrating capacity in infants andadults, taking into account the effects of dietary proteinintake on urinary concentration of urea and other solutes.

parenteral fluid therapy. Finally, the observationwas applied to infant feeding practices. Since dietswith a higher protein content result in higher ratesof excretion of urinary solutes, it was assumed thatinfants on such diets would have higher renal waterrequirements and be more susceptible to dehydra-tion during thirsting than infants receiving lowerprotein diets.These interpretations and applications all seemed

very reasonable until it was realized that none ofthem had taken into account the nature of the soluterequiring excretion and the unique handling by thekidneys of the largest single component, urea.

From observations made over 30 years ago byGamble, McKhann, Butler, and Tuthill (1934), itcould have been anticipated that during thirstingincreased excretion of urea, unlike other solutes,would require little or no additional urinary water.When their studies were recalled a few years ago andthe nature of the solute in concentrated urine was

examined, it was found, as shown in Fig. 2, thatwhereas in adults about one-half of urinary solute isurea, in infants the urine contains relatively littleurea. Extraordinarily high amounts of proteinmust be given to young infants to raise their rate ofexcretion of urea to the adult level; however, underthese circumstances, since urea binds relatively littleurinary water, it brings the concentration of totalsolutes very close to the maximum value observed inadults (Edelmann, Barnett, and Troupkou, 1960).The lower value for total osmolality in young infantsis to be explained, therefore, not primarily on thebasis of renal immaturity but almost entirely by themetabolic fact that on all ordinary diets infants storemost of the protein they ingest and excrete very littleurea.How does this new knowledge affect previous

interpretations and applications ? Physiologically,

it seems almost certain now that any lower capacitythe infant might have for concentrating solutes in theurine is due not to lack of action of antidiuretichormone but rather to differences in the mechanismsconcerned with the production and maintenance of ahigh interstitial osmolality in the renal medulla(Edelmann and Barnett, 1960). In addition toactive sodium transport, these mechanisms includethe loops of Henle functioning efficiently as counter-current multipliers and the vasa rectae as counter-current exchangers. It is known that maximuminterstitial concentration gradients are related direct-ly to the lengths of the loops of Henle, which almostcertainly are shorter in young infants (Peter, 1927).Additional work will be required, however, beforethe developmental aspects of these particular renalfunctions are understood.The situation in respect to parenteral fluid

therapy remains unchanged, since the calculation ofthe minimum volume of water needed by infants toexcrete one mOsm. urinary solute was based correct-ly, though unknowingly, almost entirely on non-urea solutes. On the other hand, since renal waterrequirements are determined mainly by the rate ofexcretion ofnon-urea solutes, which are concentratedby older children and adults also only to a maximumof about 700 mOsm./l., it should be pointed out thatarguments for the use of hypotonic intravenoussolutions apply equally well to them.

Application of these physiological principles toinfant feeding practices required considerablerevision. Contrary to previous assumptions, theprotein content of the diet has little effect either onwater conservation or on resistance to dehydrationduring episodes of complete thirsting such as mayoccur during acute infections in young infants(Edelmann and Barnett, 1960). Under circum-stances simulating the type of large extrarenal lossesof hypotonic fluid that may occur during acutediarrhoea, the water content of the diet must bereduced to about 30% of control values before thecomposition of the diet exerts any differentialeffects on water balance (Drescher, Barnett, andTroupkou, 1962). Finally, with all diets usedcommonly in practice, renal water requirements aredetermined mainly by the rate of excretion of non-urea solutes, which are increased only slightly by ahigher protein diet. Contrary to earlier conclusions,therefore, the protein content of the diet has verylittle effect on minimal renal water requirements.

Repeated instances of this sort where we have hadto modify or change completely our earlier conclu-sions should serve to demonstrate that a great dealremains to be learned about the relevance ofdevelopmental renal and electrolyte physiology.

232

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Renal Diseases in Infants and ChildrenRecent advances in the understanding of many

renal diseases in infants and children can be traced,as happens so often, to the development of new

techniques. A safe procedure for performingpercutaneous renal biopsy is providing importantnew correlations between morphological, histo-chemical, and immunochemical properties of thekidney on the one hand, and physiological functionsand clinical manifestations on the other. As a resultof such investigations, it is imperative that our

concepts concerning almost every aspect of renaldisease in infants and children be re-examined, a

truly agonizing process in a field of such historicmedical contributions and yet one so riddled withspeculation.Although the study of renal biopsies is yielding

important new information about post-streptococcalacute glomerulonephritis, the most recent stimulusfor re-examining our concepts has arisen fromanother source. From experience with adultpatients, many physicians interpret the naturalhistory of post-streptococcal acute nephritis as

shown in Fig. 3. Since most adults with chronicnephritis give no history of previous renal disease, itis assumed, with no real evidence, that their diseasebegan with an unrecognized attack of acute nephritisduring childhood. It is known from systematicstudies of the urine of children recovering fromstreptococcal infections, that as high as 95% ofchildren with acute nephritis may be unrecognizedclinically. The course of the disease in thosechildren in whom it is recognized clinically and whosurvive the extremely dangerous early complica-tions, is, as shown in Fig. 4. According to thisinterpretation there is no connexion at all betweenpost-streptococcal acute nephritis in children andchronic nephritis in adults. Complete and perma-nent recovery is shown to be virtually certain inchildren surviving the early complications. Lessthan 1% of children with acute nephritis are

indicated as progressing to chronic nephritis, and inthose cases progressive disease evolves directly fromthe acute attack, rather than following a prolongedasymptomatic latent period, as postulated withoutevidence in the first interpretation.There is some evidence in support of this second

interpretation. Follow-up studies of childrenknown to have had acute nephritis have not dis-closed significant numbers of patients developingasymptomatic latent nephritis (Hebert, 1952).Such studies, of course, do not answer the more

difficult question of whether chronic nephritis inadults is a result of unrecognized acute nephritis inchildren. Since the recovery rate is so complete

CHILDHOOD

Recovery

/ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~IAcute a

(950/o unrecognized)

IEarly DeathsHeart FailureRenal FailureHypertensive Encephalopathy

ADULT-

- Chronic

UraemiI

233

i0OD

FIG. 3.-Natural history of acute glomerulonephritis inchildren as interpreted by many physicians.

among children whose symptoms permit recognitionof the disease, however, it seems unlikely that latentand chronic nephritis would occur more frequentlyin children whose initial disease clinically was so

mild that it remained unrecognized. In addition, a

cross-sectional survey of the prevalence of abnormalAddis counts among apparently healthy 12-year-oldboys (Schlesinger, Overton, and Chase, 1956) failedto reveal a prevalence rate of latent nephritis whichwould be required if, in fact, it constituted themajor source of chronic nephritis in adults. Muchmore information is required to provide a finalanswer to this difficult but important question of thenatural history of acute nephritis in children, and itcould be obtained by proper epidemiologicalmethods. Present evidence, however, does supportthe view held by most paediatric nephrologists.Why do we harp so strongly upon these interpre-

tations ofthe natural history ofnephritis in children ?The reason is that they must provide the principlesupon which some of our most important practices

CHILDHOOD£

Recovery

Acute <1% Chronicl950/% unrecognized)

Uroeaemic

ADULTHOOD

Chronic-(aetiologyustolly unknown)

FIG. 4.-Natural history of acute glomerulonephritis inchildren as interpreted by most paediatric nephrologists.

Paediatric Nephrology

I

I

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

234

are based. In addition to the early rprompt treatment of the early coiheart failure, hypertensive encephacute renal failure, the single most imof our care of a child with acute neph]say to the parents and, when appr(child about prognosis. The interpinatural history of the disease deternafter the initial stage, one says: 'Yprobably recover, but even if he appdevelop chronic nephritis many yearschild will almost certainly recover anhe will be completely well and as wnrenal disease later in life as if he 1nephritis. He may, in fact, have gainof immunity against the rare possibsequent attack.' One need not bechild with nephritis to appreciatebetween these two statements.

Advice on other important aspectsdetermined by the interpretation chistory of acute nephritis. There arexaminations by which the progress cchild with acute nephritis can be judthese are shown diagrammaticall1

AddisCount

of

Urea

l la

II II I

I II

r- I I

<10/0 ~~~~~~I

'II

3 6 9 12 2cmth.)

TIME AFTER ONSET

FIG. 5.-Laboratory data during theglomerulonephritis in childr

Abnormal urines, which can be folloonly by Addis counts, persist on themonths and may remain abnormalyears. Glomerular filtration rate, ethe clearance of urea, if initially redLnormal usually within a week or twnot until after a month, and sometinor 3 months. The only patients inseen chronic nephritis follow what weacute nephritis were those whose cleareturn to normal after the initial epIf the clearances have remained ni

- 5 X5 E- 70

0-|#*a s

Henry- L. Barnett

ecognition and returned to normal during the first few weeks, wemplications of are entirely confident that recovery will occur andLalopathy, and advise that usual activity need no longer be restricted.iportant aspect This recommendation is in sharp contrast again withritis is what we the practices of physicians who, believing chronicopriate, to the nephritis may folow a prolonged latent periodretation of the manifested only by an abnormal urine, and assumingnines whether, that restricted activity may prevent the development'our child will of chronic nephritis, advise prolonged bed-rest. Ifears to he may such restricted activity for periods ofmonths or evenlater', or, 'your years during middle childhood is unnecessary, as wed once he does believe it is, children must not be subjected to itsnlikely to have almost immeasurably harmful effects.had never had Until recently we were relatively satisfied with oured some degree clinical care of children, despite the unansweredility of a sub- questions concerning aetiology, and most relateda parent of a research was directed toward the aetiological role ofthe difference the P-haemolytic streptococcus and the possibility

of prevention. Our experience during the past fiveof care is also years, however, has disturbed our clinical confidence,

)f the natural not in treating or assessing the prognosis in childrenre a number of with post-streptococcal acute nephritis, but inAf recovery of a establishing the diagnosis itself.Lged. Some of Because, in our experience, the development iny in Fig. 5. children of chronic nephritis following acute

nephritis was so rare, we became concerned when weo a a realized that out of42 children admitted consecutive-

ly to our service with the diagnosis ofacute nephritis,3 had failed to recover as we had expected. One ofthem, a 5 -year-old girl, had had a respiratory

Normal T infection four weeks before admission, followed in_Ranga a one week by gross haematuria and oedema. On

- admission her blood pressure was 150/90 mm. Hg.'O% Urinalysis revealed 4-plus proteinuria and innumer-

able red blood cells. Her blood urea nitrogen was41 mg./100 ml., and her antistreptolysin-0 titre,

3 ---. 1,250 units. She was oliguric for two weeks but(yr.) improved gradually. Her urea clearance, however,

rose only to the range of 60-70%. Although shecourse of acute appeared clinically well during the following year,,en. her inulin clearance at the end of it was only 15 to

20 ml. per minute. Recurrence of gross haematuriaand oedema three weeks after what appeared to

wed adequately be a typical onset of acute nephritis did attracte average for 6 our attention when she entered hospital, but wefor 2 or more have seen this often enough in children who have-stimated from recovered that even in retrospect we would haveuced returns to accepted the diagnosis. The specimen obtained byro, occasionaly percutaneous renal biopsy on the 25th day of iUnessnes only after 2 showed severe glomerulonephritis with glomerularwhom we have scarring. From her clinical course and from ae believed to be biopsy done a year later, which showed severe andLrances failed to advanced hyalinization of almost all glomeruli, we)isode (Fig. 5). had to conclude that she had severe and irreversibleLormal or have renal insufficiency. The clinical course and biopsy

Iir LWAM-O%fto-a M1.5am QL.

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Paediatric Nephrologyfindings in the other two children were similar(Edelmann, Greifer, and Barnett, 1964).What is the significance ofthese recent experiences

with children who fail to recover from what clinicallyappears to be acute nephritis ? It is only recentlythat we have collected data systematically on child-ren with acute nephritis, and so we have only animpression that there may be an increased frequencyin the incidence of severe kidney disease. Thisimpression, however, is shared by other workers inthe United States as well as in other countries, andeven the possibility of an increase in the incidencedemands investigation. An important question iswhether on the basis of this recent and relativelysmall experience we need to alter our views aboutthe prognosis in acute nephritis. I should be verysorry indeed if this occurred as a result of thisdiscussion. In none of these three children did theclearances return to normal after what appeared tobe the initial attack. We continue to be quiteconfident, therefore, that the great majority ofchildren with acute nephritis, whose clearancesremain or become normal, have an excellent progno-sis. The more troublesome question concerns thenature of the kidney disease in these three childrenand the possibility that instead of post-streptococcalacute nephritis, they had, in fact, a different andmore severe form of renal disease.Without implying any similarity, it is worth

considering here the very disturbing accounts ofso-called Balkan nephritis, an endemic form ofchronic nephropathy observed in localized areas ofYugoslavia, Rumania, and Bulgaria (Griggs andHall, 1964). In some villages the prevalence ofproteinuria has been found to be as high as 33%, incontrast to the expected figure of less than 1 %, andit is estimated that in some districts 25% of deathsare due to uraemia. Despite extensive epidemiolo-gical studies, the aetiology of this serious endemictype of nephropathy remains unknown, the mostlikely cause being some unidentified toxic factor.Taken together with our recent experiences, these

observations and the increasingly frequent reportsof the haemolytic-uraemic syndrome (Gianantonio,Vitacco, Mendilaharzu, Rutty, and Mendilaharzu,1964) raise again the question of toxicologically-induced disease. Children are being exposed to anever-increasing number of potentially harmfulagents, not only iatrogenically but also in the formof food additives, insect sprays, and other environ-mental poisons. It would not be unexpected forsuch toxic exposure to result in renal disease, theagents acting either as direct nephrotoxins or assensitizing factors, singly or as co-determinants ingenetically determined individuals. In any event,2

the warnings justify our serious concern andconstant attention.

Some Future Directions of PaediatricNephrology

What are some of the advances we can expectduring the next few years in paediatric nephrology ?Studies of developmental renal physiology and itsrelevance will focus even more, I believe, on thenewborn infant with further attempts to investigaterenal function before birth and the effects of pre-natal factors on postnatal development (McCanceand Widdowson, 1954). We can expect that stillmore disturbances of specific renal tubular functionswill be discovered, and can anticipate greater under-standing of the genetic and enzymatic mechanismsinvolved. One of the unanswered questions iswhether some of these congenital renal tubulardisorders represent developmental failures or arrests,or even the low range of distribution of normaldevelopment, rather than qualitative departuresfrom normal function. Young and Levin (Peonides,Levin, and Young, 1965) and Edelmann are study-ing developmental aspects of the renal handling ofhydrogen ions partly in the attempt to answer thisquestion in various types of renal tubular acidosis.The application of statistical and epidemiological

methods will not only define clinical problems moreprecisely and assure proper therapeutic assessments,but also by identifying associations, point the waytoward new research on mechanisms and aetiologyof renal disease. For example, Kass (1960) andothers have shown that asymptomatic pyelonephritisis far more common than realized and that it can berecognized only by a careful search for bacteriuria.The recognition and treatment of such infectionsmay serve to prevent not only a major cause ofchronic renal disease but even one of the causes ofprematurity. Thus, in addition to the obviousclinical implications of these important observations,they raise new and unanticipated questions aboutmechanisms involved in the initiation of prematureonset of labour.Of great importance also to paediatricians is

Kunin's epidemiological study (Kunin, Zacha, andPaquin, 1962) showing a prevalence rate of bacteri-uria of 1 2% in schoolgirls with covert and overtincidence rates of0 - 7 and 2 * 9% per year, respective-ly. These observations raise the very practicalquestion of whether current practices are keepingpace with the acquisition of new knowledge. Theycertainly suggest that screening for bacteriuriashould be a routine procedure in paediatric practiceand in community school health programmes. If itwould be difficult if not impossible to add this

235

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Henry L. Barnettexamination to those being done routinely, thequestion must be asked whether it should not replaceone of the current examinations. Despite lack ofsuccess thus far, it can be anticipated that as theimportance of detecting bacteriuria becomes widelyappreciated, simpler methods for it will be devised.The point is, however, that clinical practices, likeconcepts of disease, need continuously to be re-examined.The major unanswered questions concerning the

treatment of children with the nephrotic syndromeprovide striking examples of both the need for andthe neglect of the application of statistical methods.When adrenocortical steroids were first used over 15years ago, no attempt was made to assess properlytheir therapeutic value by a controlled trial.Unfortunately, as so often happens, the criticalperiod during which it could have been doneethically was lost. Most paediatricians believe nowthat, in addition to controlling oedema, theseagents have a favourable effect on the underlyingrenal disease. On the other hand, unlike the rareevent in which a drug like penicillin is almostcompletely effective against certain infections, thevalue of adrenocortical steroid therapy in childrenwith the nephrotic syndrome is not so unequivocalthat its universal use can be accepted withoutquestion, especially when due attention is paid to thetoxic side-effects. Differences between the majorquestions being asked in two clinical trials beingconducted in this country illustrate these points.There is enough uncertainty about the value ofadrenocortical steroids in the treatment of adultswith the nephrotic syndrome that the trial beingconducted by the Medical Research Council has acontrol group receiving no steroids. In contrast,the collaborative clinical trial in children being co-ordinated by members of the Department of SocialMedicine at St. Thomas' Hospital and of theDepartment of Paediatrics at Guy's Hospital canask only the interesting but less important questionof whether one of two steroids given by one of tworegimens is more effective. Both trials have thegreat advantage that the results are being correlatedwith examination of tissue obtained by percutaneousrenal biopsy. It should be of interest to paediatri-cians that the basic question of whether or notadrenocortical steroids are actually indicated inadults with the nephrotic syndrome is still open andthat the answer may depend in part upon the type ofglomerular pathology. From present informationit would not be ethical to withhold steroids from agroup of children with the nephrotic syndrome forthe purpose of therapeutic assessment. On theother hand, if additional correlations between the

clinical course of the disease in both children andadults with various types of glomerular pathologyshould permit identification of groups of childrenwho are very likely either to recover without steroidsor to fail to respond to such therapy, a controlledtherapeutic trial may still be indicated to assess thebalance between therapeutic value and toxicity.A more immediate decision of this type may be at

hand. Constantly increasing experimental andclinical evidence supports the hypothesis thatimmunological mechanisms play a central role in thepathogenesis of the nephrotic syndrome. Therecent demonstration by Heymann, Hunter, Hackel,and Cuppage (1962) of an autoimmune renaldisease in rats adds important experimental supportfor this concept. Using immunofluorescent tech-niques and electronmicroscopy, Vernier and associ-ates (Michael, Drummond, Good, and Vernier,1964) have shown that some children with thenephrotic syndrome have deposits of y and 3-l-cglobulin on the glomerular basement membrane,and that the presence of these immunoglobulins isusually associated clinically with resistance toadrenocortical steroids. These patients appear torespond to immuno-suppressive agents as havesome of our patients who, though responsive tosteroids, developed such severe toxic effects that thedoses required to suppress their proteinuria couldnot be continued. We are at the stage in the treat-ment of this group of patients where we areuncertain about the value oftreatment with immuno-suppressive agents and where not only is a propertherapeutic trial justified but also failure to do onecould almost be considered unethical, as suggestedby Sir Austin Bradford Hill (1952) when he said:'It may sometimes be unethical not to experiment.'Since the number of such patients is small, acollaborative trial with its attendant difficultieswould be required. Nevertheless, it can be hopedthat the experience of having learned so much lessthan we should have from the uncontrolled use ofadrenocortical steroids in children with the nephroticsyndrome will persuade us not to miss the presentcritical period during which the value of thesetherapeutic agents can be assessed properly.

SummaryTo sum up, then, I hope I have demonstrated by

example that the scope, complexity, and importanceof the problems involved require a speciality ofpaediatric nephrology. Contrary to the deprecatingand erroneous statement that scientific specializationinvolves knowing more and more about less and less,the paediatric nephrologist must draw upon know-ledge from an ever-increasing number of disciplines.

236

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from

Paediatric Nephrology 237He must have sound basic paediatric training sincehis general knowledge of infants and children and ofhuman development is what distinguishes him fromothers whose major interests are in the kidneys andtheir diseases. Without paediatric training hecould not apply information on the development ofrenal function in young infants to the problems offluid and electrolyte therapy, nor would he be likelyto understand how the interpretation of the naturalhistory of nephritis in children affects his discussionsof prognosis with parents. In addition to paedia-trics, he must have extensive training in at least oneand some knowledge of many other disciplines,including renal physiology and pathology, immuno-logy, endocrinology, enzymology, genetics, andepidemiology. As a clinical investigator he mustserve a 'bridging' role, doing person-orientedresearch, often in co-operation with pre-clinicalscientists, on the one hand, and specialized clinicalconsultation with general paediatricians or familyphysicians, on the other.

Increasing specialization of the type I amadvocating affects not only paediatric research butalso the teaching and practice of paediatrics. Someof the difficult problems raised are important to thewhole future of paediatrics. However these aresolved, and there are probably several differentsolutions, I do not believe the cause of paediatricswill be served if, in the process, specialities such aspaediatric nephrology cannot flourish.

This lecture was prepared during a sabbatical leavesupported in part by a fellowship from the Common-wealth Fund of New York while enjoying the hospitalityof the Department of Medical Statistics and Epidemiolo-gy of the London School of Hygiene and TropicalMedicine. Some of the investigations referred to weresupported by grants from the National Heart Institute,the National Institute of Child Health and HumanDevelopment, The Division of Research Facilities andResources, the Health Research Council of The City ofNew York, the Kidney Disease Foundation of NewYork, and the Sylvan League, Inc.

REFERENCESBarnett, H. L. (1940). Renal physiology in infants and children: I.

Method for estimation of glomerular filtration rate. Proc. Soc.exp. Biol. (N. Y.), 44, 654.

-, McNamara, H., Shultz, S., and Tompsett, R. (1949). Renalclearances of sodium penicillin G, procaine penicillin G, andinulin in infants and children. Pediatrics, 3, 418.

Bedson, S. P. (1943). Some recent virus work and its practicalimport. Arch. Dis. Childh., 18, 113.

Crewe, F. A. E. (1955). Medicine and social science. ibid., 30, 303.Drescher, A. N., Barnett, H. L., and Troupkou, V. (1962). Water

balance in infants during water deprivation. Amer. J. Dis.Child., 104, 366.

Edelmann, C. M., Jr., and Barnett, H. L. (1960). Role of thekidney in water metabolism in young infants. J. Pediat., 56,154.

-, -, and Troupkou, V. (1960). Renal concentrating mech-anisms in newborn infants. Effect of dietary protein andwater content, role of urea, and responsiveness to antidiuretichormone. J. clin. Invest., 39, 1062.

-, Greifer, I., and Barnett, H. L. (1964). The nature of kidneydisease in children who fail to recover from apparent acuteglomerulonephritis. J. Pediat., 64, 879.

Finberg, L., Kiley, J., and Luttrell, C. N. (1963). Mass accidentalsalt poisoning in infancy. J. Amer. med. Ass., 184, 187.

Gamble, J. L., McKhann, C. F., Butler, A. M., and Tuthill, E.(1934). An economy of water in renal function referable tourea. Amer. J. Physiol., 109, 139.

Gianantonio, C., Vitacco, M., Mendilaharzu, F., Rutty, A., andMendilaharzu, J. (1964). The hemolytic-uremic syndrome.J. Pediat., 64, 478.

Griggs, R. C., and Hall, P. W. (1964). Investigations of chronicendemic nephropathy in Yugoslavia. In Renal Metabolism andEpidemiology of Some Renal Diseases. Proc., 15th AnnualConference on The Kidney, ed. J. Metcoff, p. 312. NationalKidney Foundation, New York.

Hebert, H. J. (1952). Acute glomerulonephritis in childhood. Astudy of the late prognosis of twenty-seven cases. J. Pediat.,40, 549.

Heymann, W., Hunter, J. L. P., Hackel, D. B., and Cuppage, F.(1962). Transfer of experimental auto-immune nephrosis inrats. Proc. Soc. exp. Biol. (N.Y.), 111, 568.

Hill, A. B. (1952). The clinical trial. New Engl. j. Med., 247, 113.Kass, E. H. (1960). The role of asymptomatic bacteriuria in the

pathogenesis of pyelonephritis. In Biology of Pyelonephritis,ed. E. L. Quinn and E. H. Kass, p. 399. Little, Brown,Boston.

Kunin, C. M., Zacha, E., and Paquin, A. J., Jr. (1962). Urinary-tract infections in schoolchildren. I. Prevalence of bacteriuriaand associated urologic findings. New Engl.J. Med., 266, 1287.

McCance, R. A., and Widdowson, E. M. (1954). The influence ofevents during the last few days in utero on tissue destruction andrenal function in the first two days of independent life. Arch.Dis. Childh., 29, 495.

Michael, A. F., Drummond, K. N., Good, R. A., and Vernier, R. L.(1964). Immunoglobulins: clarification of their significance inrenal disease and demonstration of response to immuno-suppressive therapy. .T- clin. Invest., 43, 1291.

Peonides, A., Levin, B., and Young, W. F. (1965). The renalexcretion of hydrogen ions in infants and children. Arch. Dis.Childh., 40, 33.

Peter, K. (1927). Untersuchungen uber Bau und Entwicklung derNiere. G. Fischer, Jena.

Schlesinger, E. R., Overton, D. E., and Chase, H. C. (1956). Studyof children drinking fluoridated and nonfluoridated water.Quantitative urinary excretion of albumin and formed elements.J. Amer. med. Ass., 160, 21.

Sheldon, W. (1956). Pediatric profiles, George Frederic Still.J. Pediat., 49, 229.

copyright. on F

ebruary 16, 2021 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.41.217.229 on 1 June 1966. Dow

nloaded from