Journal of Clinical InvestigationVol. 45, No. 4, 1966

The Nephrotic Syndrome of Childhood: Immunologic,Clinical, and Pathologic Correlations *

KEITH N. DRUMMOND,t ALFREDF. MIICHAEL,4 ROBERTA. GOOD,§ ANDROBERTL. VERNIER

(From the Pediatric Research Laboratories of the Variety Club Heart Hospital, and theDepartment of Microbiology, University of Minnesota, Minneapolis, Minn., and

the Department of Pediatrics, McGill University, Montreal, Canada)

The role of immunologic processes in the patho-genesis of human renal disease has been the sub-ject of investigation and discussion for over half acentury. Factors interpreted as favoring an im-munologic basis for some of the diffuse glomerulardiseases of man include reduced serum comple-ment levels, the presence of antikidney antibodies,and the similarity of pathologic changes in im-munologically induced experimental renal diseaseto lesions in human glomerulonephritis. Morerecently, the demonstration by immunofluorescenttechniques of glomerular deposition of immuno-globulins and complement in a number of thesedisorders has provided more compelling supportfor the immunologic hypothesis. *The possiblepathogenetic mechanisms by which immunologicprocesses may induce renal disease in man havebeen recently reviewed (2).

Most immunofluorescent studies of kidney biop-sies in the nephrotic syndrome have suggested thatimmunoglobulin deposition along glomerular base-ment membranes is characteristic (3-7). Suchstudies have dealt predominantly with adult ne-phrotic patients, a group in which membranousglomerulonephritis is a frequent pathologic alter-

* Submitted for publication September 2, 1965; acceptedJanuary 3, 1966.

Aided by grants from the U. S. Public Health Service(AI-02018, HE-05662, HE-06314), the National Founda-tion, The American Heart Association, and the MedicalResearch Council of Canada (MA-1579).

Presented in part at the Annual Meeting of the Societyfor Pediatric Research, Seattle, Wash., June 20, 1964 (1).

t Address requests for reprints to Dr. Keith N. Drum-mond, the Montreal Children's Hospital, 2300 Tupper St.,Montreal 25, Quebec, Canada.

t Established Investigator, American Heart Association.§ American Legion Memorial Heart Research Pro-

fessor of Pediatrics and Microbiology.11 Present address: Dept. of Pediatrics, University of

California, Los Angeles, Calif.

ation. However, even in patients with the idio-pathic nephrotic syndrome, positive staining forgamma globulin and complement has been de-scribed as a characteristic finding (6). Similarobservations have been made in infants with thecongenital nephrotic syndrome (8, 9).

The purpose of this work was to obtain moredefinitive information about the significance of im-munologic processes in a relatively large group ofpatients with the nephrotic syndrome of childhood.Our observations were carried out in 35 childrenwith this disease. Deposition of immunoglobulinG (IgG) and beta1c-globulin in glomeruli wasstudied by the imunofluorescent technique and cor-related with light microscopic findings and with theclinical course and response to therapy. Serumcomplement activity was measured in 10 of thesepatients. Wefound that in children whose neph-rotic syndrome is responsive to steroid therapyand whose renal pathology is minimal, immuno-globulin and complement are not demonstrable onthe glomerular basement membrane, suggestingthat immunologic processes may not be of patho-genetic importance in this group. However, suchmechanisms do appear to be operative in many ofthe nephrotic children whose clinical courses andkidney lesions are indicative of severe glomerularinjury.

Methods

Subjects. The patient group consisted of 35 childrenwho were admitted to the University of Minnesota Hos-pitals between 1962 and 1964 with the clinical and labora-tory features of the nephrotic syndrome. Patients wereconsidered to have this disorder if they presented the fol-lowing features: 1) generalized edema, with or withoutascites; 2) marked proteinuria as judged by a urinaryprotein concentration of 1,000 mg per 100 ml or more, a24-hour urinary protein excretion of 2 g per m2 or more,or both; 3) depression of serum albumin below 2 g per100 ml; and 4) elevation of serum cholesterol above

620

NEPHROTICSYNDROMEOF CHILDHOOD

FIG. 1. NOFLUORESCENCEIS SEEN IN THESE TWOGLO-MERULI FROMA PATIENT IN GROUPI, SUBGROUPA, STAINEDFOR IMMUNOGLOBULiNG (TGG). (x 200.)

2.0 mg per 100 ml. In most patients all features were

present at the time of the study biopsy. However, in a

few from whom biopsies were taken early in the course

of an exacerbation, after onset of improvement, or afterthe development of chronic renal disease, not all featureswere present at the actual time of biopsy.

One patient had a sore throat 1 month preceding on-

set of the nephrotic syndrome and showed an antistrep-tolysin 0 titer of 500 Todd U at the time of biopsy.The other patients had no clinical or laboratory evidenceof preceding streptococcal infection. None had gross he-

maturia, and in none was the diagnosis of acute post-streptococcal glomerulonephritis considered likely.

Two patients with the nephrotic syndrome developingin the course of systemic lupus erythematosus were notincluded in the study. Laboratory, clinical, and pathologicfindings were considered adequate to exclude this diag-nosis in the remaining patients.

In one patient it was possible to relate the develop-ment of the nephrotic syndrome to preceding Tridione1administration. In the others, a distinctive disease proc-ess or known precipitating cause was not apparent.

Tissue preparation. Renal tissue was obtained by per-cutaneous biopsy or at autopsy and was divided into twopieces. One portion was fixed in 10% buffered formalin

1 Trimethadione, Abbott Laboratories, North Chicago,Ill.

FIG. 2. A) THIS GLOMERULUSFROMANOTHERPATIENTIN GROUP I, SUBGROUPA, STAINED FOR BETAiC-GLOBULINSHOWSNO FLUORESCENCE. (X 400.) B) LIGHT MICROS-COPY OF A GLOMERULUSFROMTHE SAMEPATIENT REVEALSNO ABNORMALITY. (PAS X 400.) PAS= periodic acidSchiff stain.

621

DRUMMOND,MICHAEL, GOOD, AND VERNIER

(pH 7.35) and processed for routine pathologic examina-tion with the periodic acid Schiff (PAS), hematoxylinand eosin (H and E), and azocarmine stains on 4-es sec-tions. The other piece was immediately frozen in iso-pentane that had been precooled to - 1700 C by liquidnitrogen. This tissue was kept frozen at - 70° C untilsectioned at 4 tt on a Lipshaw cryotome at - 200 C. Thetissue was stained by the direct immunofluorescent methodwith fluorescein isothiocyanate (FITC) -labeled antisera.Antisera were prepared in rabbits to purified human IgG,2albumin,3 and betaic-globulin,4 and in a goat to rabbitgamma globulin.5 The methods used for immunization,antisera preparation, FITC conjugation, tissue sectioning,staining, and microscopic and photographic techniques aredescribed in another publication (11). Antisera to hu-man albumin and to rabbit gammaglobulin were used ascontrol stains for the antisera to human IgG and betac-globulin. In no instance was positive glomerular stain-ing noted with these control stains. Specificity of IgGand betaic-globulin fluorescence was established by usingan inhibition test in which the section was pretreatedwith unlabeled antisera. These sections were then com-pared with those in which the inhibition step with un-labeled antisera had been omitted. Reduction in fluores-cence in the inhibited sections was interpreted as indi-cating specificity for IgG or beta1c-globulin fluorescenceand was uniformly seen in all sections so tested. Thetissues were examined independently by two investigatorsusing a Zeiss fluorescent microscope.

Serum complement activity was determined by themethod of Kabat and Mayer (12) 6 in 10 of the pa-tients within 1 week of the biopsy. In this method thelevel of complement activity is expressed as the titer ofserum causing 50% hemolysis of a test suspension ofsensitized sheep erythrocytes. Control values in ourlaboratory are 45 + 8 (mean + 1 SD).

Sections for routine light microscopy were examinedindependently by at least two of the authors and by apathologist who had no knowledge of the clinical status

2 Isolated by DEAE chromatography; IgG was alsoobtained from Immunology, Inc., Chicago, Ill.

3 Normal human serum albumin, Cutter Laboratories,Berkeley, Calif.

4 Prepared according to the method of Mfiller-Eberhard,Nilsson, and Aronsson (10).

5Rabbit gammaglobulin, Cohn Fraction II, NutritionalBiochemicals Corp., Cleveland, Ohio.

6These determinations were performed by Dr. HenryGewurz of the Dept. of Pediatrics, University of Minne-sota.

TIENT STAINED FOR BETAie-GLOBULIN. This glomerulusshows more extensive fluorescence, which arises in theregion of the afferent and efferent arterioles and appears

FIG. 3. A) ONLY THE REGION OF THE AFFERENT AR- to extend into the stalk or mesangium. The exact locali-TERIOLE IS FLUORESCENTIN THIS GLOMERULUSFROMA PA- zation of the fluorescence cannot be determined; however,TIENT IN GROUP II, SUBGROUPD, STAINED FOR IGG. it is clearly not along the capillary basement membranes.(X 400.) B) ANOTHERGLOMERULUSFROMTHE SAMEPA- (X 400.)

622

NEPHROTICSYNDROMEOF CHILDHOOD

FIG. 4. A) TWOGLOMERULI FROM THE PATIENT IN

GROUP III WITH MEMBRANOUSGLOMERULONEPHRITIS

SHOWINGMARKEDDEPOSITION OF IGG ALONG THE CAPIL-

LARY BASEMENT MEMBRANES. (X 250.) B) DIFFUSE

of the patient.7 An over-all assessment of the degree ofabnormality was made; in addition, the specific abnor-mal features were graded on a scale ranging from 1 +(minimal) to 4 + (marked).

Results

Immunofluorescent studiesOn the basis of the immunofluorescent findings

it was possible to separate the patients into threegroups. The first group (I) consisted of 22 pa-tients whose renal glomeruli showed no depositionof IgG or betalc-globulin (Figures 1, 2A).

In the second group (II) were seven patientswhose glomeruli showed a coarse focal or basaldistribution of IgG or betaic-globulin. Depositionof these proteins along glomerular basement mem-branes was negligible. Occasionally the fluores-cence was confined to the stalk or mesangial re-gion or to the basal area near the afferent or ef-ferent arterioles. In some instances the fluores-cence appeared to be within capillary lumina or inBowman's space. Often the area of fluorescencewas only one-tenth to one-twentieth of the glo-merular surface area, and different glomeruli froma given specimen were involved to a variable de-gree (Figure 3, A and B).

The third group (III) comprised six patientswhose biopsies showed extensive glomerular depo-sition of both IgG and beta1c-globulin. The depo-sition was usually linear and appeared to be onthe basement membrane, although in some casesnodular excrescences and lumpy deposits along thebasement membrane were also observed. Re-gardless of the predominant pattern it was clearthat the fluorescence was related to the basementmembranes of the capillary loops of the glomeruli(Figure 4A).

Clinical course and routine laboratory studies(Tables I and II)Group I. These 22 patients showing no glo-

merular fluorescence could be separated into twogroups, A and B, on the basis of their responses to

7 We are indebted to Dr. Barbara Burke of the De-partments of Pediatrics and Pathology, University ofMinnesota, for her assistance in this evaluation.

THICKENING OF THE CAPILLARY BASEMENTMEMBRANESISSEEN IN THE LIGHT MICROSCOPICSTUDY OF THE SAMEPA-TIENT. (PAS X 400.)

623

DRUMMOND,MICHAEL, GOOD, ANDVERNIER

TABLE I

Clinical features and laboratory

No. ofpreceding Clinical status

Group Patient Sex Age at onsett Age at biopsy episodes at biopsy:

---.L-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~moyears

IA 1 M 62 M 33 F 54 M 55 M 46 M 57 M 18 M 29 M 2

10 M 111 F 312 M 113 M 114 F 1415 MI 816 M 2

IB 171819202122

HIC

IID

M 3I

M 1M 1F 5F 3

nmns years

3 66 105 82 5

115

8 176 5

26 10

83 46 68 144 82 2

months

4113

15

10651

1010

963

0 1 + edema1 4+ edema, ascites4 1 + edema, ascites0 1+ edema8 1 + edema, ascites0 3 + edema, ascites8 1 + edema7 1 + edema2 3 + edema, ascites

18 3+ edema, ascites3 No edema1 No edema8 No edema0 1 + edema0 3 + edema0 2+ edema, ascites

1 3 3 0 , 1+ edema5 6 0 4+ edema, ascites1 1 10 0 4+ edema, ascites1 1 3 0 4+ edema, ascites3 13 5 7 3+ edema, ascites

6 3 6 2+ edema, anemia,azotemia, hyper-tension

23 F 2 10 5 7 1 No edema24 F 2 8 8 9 1+ edema, ascites25 M 9 7 9 10 1 3+ edema, ascites26 M 5 8 6 5 1 1+ edema

27 F 1428 F 2

29 MI 7

14 4 0 1+ edema10 2 11 0 Tridione for 1 year

4+ edema, ascites8 7 10 0 3+ edema, ascites

III 30 F31 F32 MI33 M

34 M

35 F

12 2 12 3 0 3+ edema, ascites11 3 11 4 0 1+ edema

5 6 5 7 0 3+ edema, ascites12 20 8 2+ edema, uremia,

anemia7 10 9 1 2 Hypertension-no

edema3 6 8 3 1+ edema, ascites,

hypertension

* Abbreviations: BUN= blood urea nitrogen, Ccr = creatinine clearance.t The age at which the first episode of the nephrotic syndrome occurred.4 The degree of edema and ascites has been graded from 1 + (minimal) to 4+ (marked).§ C'H5o represents the titer of serum that will cause 50% lysis of a test suspension of sensitized sheep erythrocytes.

Normal control values in our laboratory are 45 8 (mean ± 1 SD).

steroid therapy. In this and the subsequent groupspatients were considered to have responded to ster-oid therapy or to have had cessation of proteinuriaif the qualitative test for urinary protein was nega-tive or if the excretion of protein was less than100 mg per m2 per 24 hours. Patients were con-sidered resistant to steroid therapy if the featuresof the nephrotic syndrome, as described earlier,

persisted after 4 to 6 weeks of treatment (predni-sone, 60 mg per m2 per day).

Subgroup A. The 16 patients in this subgroupwere in most respects typical of children with theidiopathic nephrotic syndrome. Although the ageof onset ranged from 1 year 3 months to 14 years8 months, in only three instances was it over 6years. Ten of the patients had had previous epi-

624

NEPHROTICSYNDROMEOF CHILDHOOD

TABLE I

data at time of study biopsy*

Serum proteins Serumcholes-

Albumin Globulin terol Urinary protein BUN Cc. C'H6o§ Response I

g/100 ml mg/100 ml

1.1 3.4 5420.7 3.2 7560.7 3.0 5280.7 2.9 5700.9 2.9 4760.7 3.2 620

2471.0 3.7 4090.8 2.7 4141.8 2.7 3371.7 2.8 4062.6 2.5 2272.7 3.0 1771.5 2.9 4980.5 3.7 4870.4 3.0 353

0.8 2.4 3780.9 2.9 2640.3 3.8 6420.2 2.1 3550.6 2.9 5420.9 3.5 722

g/24 hours

3.24+S 3.53+ 2.1

1.74+ 19.03+ 1.33+ 2.84+ 3.64+ 1.13+ 6.04+ 2.24+ 2.42+ 0.4

20.03.8

1+ 2.5

3+ 1.94+ 5.04+ 8.04+ 17.04+ 5.72+ 6.0

mg/lOO ml

7

1218

8168

1314

11151216

813

17261258

7226

L/24 hours/1.73 m'

110125131150

154177

178

136129

X 200107

16948

156

+40- +

++

+++42 +38 +

+

++

43 +

33 -29 -

_**

2.1 3.3 96 1 + 0.30.5 3.3 375 3+ 2.70.4 2.8 532 4+1.6 3.2 297 4+ 1.5

1.8 2.7 4061.6 2.5 391

1.0 2.9 406

4+

3+

6.613.0

2099

+33 +

86 +220 34 +

10 14225 33

5.2 20 76

1.0 2.2 274 3+ 8.0 640.7 2.6 528 3+ 2.7 750.5 3.4 440 4+ 3.9 881.8 2.2 279 4+ 2.6 190

1.5 2.3 357 4+ 6.4 21

1.6

34 40 -20

_ **

87 14 -

1.9 440 3+ 6.2 14 107

11 Refers to the 4- to 6-week period after onset of therapy; + = cessation of proteinuria within 1 month; - = failureto respond.

¶ Qualitative test for urinary protein: 1 + = 30 to 100 mg per 100 ml, 2 + = 100 to 500 mg per 100 ml, 3 + = 500 to1,000 mg per 100 ml, and 4+ = more than 1,000 mg per 100 ml.

** Deceased.

sodes of the nephrotic syndrome, and all had ex-perienced cessation of proteinuria with previoustherapy. At the time of biopsy, edema, ascites, orboth, were present in all but three patients, al-though each of these had shown these physicalfindings in the past. None of the patients werereceiving daily steroid therapy at the time ofbiopsy.

Marked proteinuria was present in all patients,and depression of the serum albumin was seen inall but one, in whom it was not measured, al-though during this patient's previous nephroticepisodes hypoalbuminemia was documented. Se-rum cholesterol values were elevated in all but onepatient, who had recently responded to steroidtherapy and had protein-free urine until 3 days

625

DRUMMOND,MICHAEL, GOOD, ANDVERNIER

TABLE IICorrelation of immunofluorescent, morphologic, and clinical characteristics in 35 children with the nephrotic syndrome

Renal pathology

Mini-mal Serum

Deposition of abnor- complementIgG* and betaic- No. of Response to steroid Nor- mali- Abnor- levelst

Group globulin Patients therapyt mal ties mal Comments

I Negative 22 A. Responsive 16 2 13 Normal 4 Patients with subsequent exacerbationshave remained steroid responsive.

B. Resistant 6 2 4 Normal 2 Three dead with renal failure.

II Coarse, focal, 7 C. Responsive 4 4 Normal 2 No clinical or laboratory differencesstalk, or basal from group IA.

D. Resistant 3 2 1 Tridione§ administration precedingnephrotic syndrome in one patient.

III Basement 6 Resistant 6 6 Normal 1 One dead with renal failure.membrane Low ill

* Immunoglobulin G.Refers to the 4- to 16-week period after onset of therapy. A, B, C, and D are subgroups.All serum complement levels were obtained within 7, and in most instances within 4, days of study biopsy.

¶ Trimethadione.Greater than 2 SD below normal control values.

before biopsy. Normal values for creatinine clear-ance were found in the 11 patients in whom it was

determined. These 16 patients were characterizedby the responsiveness of their nephrotic syndrometo steroid therapy. Within 20 days of initiatingdaily steroid therapy (prednisone, 60 mg per m2per day) all experienced cessation of proteinuria,and when edema was present a diuresis occurred.In those patients who have had subsequent ex-

acerbations a continuing responsiveness to steroidtherapy has been noted.

Subgroup B. The second group of patients withno glomerular fluorescence consisted of six pa-

tients who failed to respond to steroid therapy.One patient was only 5, and two were only 13,months of age when the nephrotic syndrome was

diagnosed; otherwise, the clinical features thatthese children presented did not distinguish themfrom those of subgroup A. The laboratory stud-ies in these six patients revealed essentially thesame features as seen in those of subgroup A, ex-

cept that three patients had evidence of glomerularinsufficiency, as indicated by elevation of the bloodurea nitrogen, reduced creatinine clearance, or

both.These patients were resistant to what were con-

sidered adequate trials of steroid therapy-pred-nisone, 60 mg per m2 per day for 1 to 2 months.In three patients, death ultimately resulted fromthe renal disease. In the others, persistent pro-

teinuria and other clinical and laboratory findingsindicated continued active 'renal disease.

Group II. This group of seven patients was

characterized by a coarse focal pattern of glomeru-lar IgG and betaic-globulin deposition. These pa-tients could also be subdivided into two subgroups(C and D) on the basis of their responses to ster-oid therapy.

Subgroup C. This consisted of four patientswho had experienced at least one previous epi-sode of the nephrotic syndrome from which theyhad recovered with cessation of proteinuria. Onthe basis of their clinical and routine labora-tory findings and their favorable responses to ster-oid therapy, it was not possible to separate thesepatients from those of Group I (subgroup A) whoresponded to steroid therapy.

Subgroup D. This subgroup consisted of threepatients who failed to respond to steroid therapy.One had taken Tridione as treatment for seizuresin the year preceding the clinical onset of the ne-phrotic syndrome. Two had reduction of creati-nine clearance and elevation of blood urea nitrogenat the time of the study. Otherwise, these patientswere not separable on the basis of their clinical andlaboratory findings from the other patients ofgroups I and II.

Group III. This group of six nephrotic patientsshowing glomerular basement membrane deposi-tion of IgG and betaic-globulin differed in severalrespects from the patients in groups I and II.Their average age was higher (three patients wereover 11 years old at the time of onset), and in gen-eral their clinical and laboratory findings indicatedmore severe disease than in the patients of groupsI and II. Each patient showed microscopic he-

626

NEPHROTICSYNDROMEOF CHILDHOOD

maturia, five had reduction in glomerular filtrationrate, several were hypertensive, and one who hassince died was in advanced renal insufficiency withuremic pericarditis. None responded to the usualcourses of steroid therapy.

Renal pathology of group I (Figure 2B)Subgroup A. The renal tissue in two of the

patients showed no abnormality. In one, althougha suitable specimen was obtained for immuno-fluorescent study, no glomeruli were present inthe routine pathologic preparation. Minimal ab-normalities were seen in the other 13, includingfocal glomerular hypercellularity, segmentation ofthe glomerular tuft, prominence or proliferationof the mesangial or stalk region, focal thickeningof the glomerular basement membranes, foci of in-terstitial calcium deposition, and tubular dilatationor atrophy. Six patients showed infrequent hy-alinized glomeruli. These abnormalities were notstriking; when judged on a scale of severity from1 + (minimal) to 4 + (marked), the great ma-

jority were considered to be 1 + and none were

worse than 2 +.Subgroup B. By contrast, of the six patients

with no glomerular fluorescence who resisted ster-oid therapy, four showed clearly abnormal renaltissue. In each of these patients there was markedmesangial or stalk proliferation. In addition, fo-cal glomerular scarring, interstitial fibrosis, hy-alinized glomeruli, and tubular dilatation were

seen in some of them. Two showed minimalabnormalities similar in type and severity to thoseseen in patients who responded to steroid therapy.

Renal pathology of group II

Renal abnormalities in this group were similarto those of group I. In each of the patients whoresponded to steroid therapy (subgroup C) onlyminimal abnormalities, indistinguishable fromthose of group I, subgroup A, were seen. Of thesteroid resistant patients (subgroup D), two hadminimal abnormalities, and one showed clearly ab-normal renal tissue with epithelial crescents andcapsular adhesions.

Renal pathology of group III

All patients in this group showed severe patho-logic changes. In three, the lesions of prolifera-tive subacute glomerulonephritis were found. In

two the lesions, although as severe, were morechronic, in that scarring was more pronounced andbasement membrane thickening more widespread;these changes are compatible with the pathologicdiagnosis of chronic glomerulonephritis. In onepatient, the changes were those of a diffuse mem-branous glomerulonephritis (Figure 4B), a rela-tively common pathologic finding in adults withthe nephrotic syndrome (13), but uncommon innephrotic children.

Serum complement levels. Each of the eightvalues obtained from the patients in group I andII fell within normal limits; in group III one valuewas normal and one was low.

DiscussionThese data suggest that it is possible, on the

basis of immunopathologic studies, to delineatetwo groups of children with nephrotic syndrome.In one, comprising the majority of children withnephrosis, no evidence that immunologic mecha-nisms are operative was obtained with the methodsused; in the other, immunologic mechanisms areprobably of pathogenetic importance. The lattergroup, represented by our group III, was sepa-rable on the basis of clinical, laboratory, or patho-logic findings from patients of groups I and II.The diagnoses include membranous glomerulo-nephritis [or membranous nephrosis of Churg andassociates (13)] and subacute and chronic glo-merulonephritis. Deposition of IgG and betalc-globulin was seen along the glomerular basementmembranes. These patients all failed to respondto conventional doses of corticosteroids, but sev-eral did respond to subsequent large doses of im-munosuppressive agents, among them prednisone(14). Group III constitutes only a relativelysmall proportion of children with the nephroticsyndrome. However, since the nephrotic syn-drome can occur as a manifestation of acute post-streptococal glomerulonephritis (15), and sincerecent studies have established that immunologicmechanisms are operative in this disease (11, 16,17), patients developing the nephrotic syndromein the course of this disease could be included inthis group.

Patients in groups I and II satisfy the clinicaland pathologic criteria of idiopathic nephroticsyndrome of childhood. In none of these patientswas IgG or betalc-globulin detected along the glo-merular basement membrane. In group I all glo-

627

DRUMMOND,MICHAEL, GOOD, AND VERNIER

meruli examined were completely negative, whereasin group II there were focal glomerular deposits ofthese immunoproteins. These appeared to liewithin the mesangium, at the base of the glomeru-lus in the region of the afferent or efferent arteri-oles, within capillary lumina, or within Bowman'sspace. The significance of these immune depositsis not clear; however, we are inclined to interpretthem as secondary to the proteinuria and not ofprimary pathogenetic significance since they arenot localized in the glomerular membranes at thesite of the functional lesion. Further study is nec-essary to clarify these relationships, but certainlythe quantity and distribution of the immunoglobu-lin and complement deposits in group II clearlydistinguish them from those of group III, whichinvolved most of the glomerular surface of all theglomeruli examined and were located predomi-nantly along the capillary basement membrane.

In view of previously published reports of glo-merular immunoprotein deposition as a charac-teristic finding in the idiopathic nephrotic syn-drome (3-7), it is important to attempt reconcili-ation of such observations with the results of thepresent study. First, most of these reports havedealt with adults. In adults, membranous glo-merulonephritis [or membranous nephrosis (13) ]is a common pathologic finding. By contrast,children with nephrosis rarely have diffuse thick-ening of the basement membrane (18). In ourone patient with membranous glomerulonephritis,deposition of IgG and betaic-globulin was foundon the glomerular basement membranes. Sec-ondly, at low intensities of fluorescence, subjectiveand technical factors become more important; wemay be underinterpreting, or others overinter-preting, the results obtained. The question is asignificant one, nevertheless, and here certain ob-servations are pertinent. In patients with earlydisseminated lupus erythematosus in whom renaldisease is not detectable by clinical and laboratoryevaluation, minimal deposits of IgG and betalc-globulin are readily detectable along glomerularbasement membranes (19). In patients with acutepoststreptococcal glomerulonephritis showing min-imal proteinuria, deposits of these immunoproteinsare easily found along capillary basement mem-branes (11). When massive proteinuria occursin these diseases, extensive deposition of IgG andbeta1c-globulin is always present along the glo-

merular basement membranes. Lupus nephritisand acute poststreptococcal glomerulonephritis areconsidered the prototypes of human renal diseasesthat have an immunologic basis. It seems ex-tremely unlikely that immunoprotein deposits onbasement membranes large enough to be of patho-genetic importance in the development of massiveproteinuria of the children of groups I and II areescaping detection by the techniques employed inthis study. Moreover, in immunologically inducedexperimental renal disease in animals [nephrotoxicserum nephritis (20), antigen-antibody complexnephritis (21), and the "autoimmune" model ofHeymann (22, 23)], renal damage sufficient toresult in proteinuria is associated with readily de-tectable deposits of immunoproteins on glomeru-lar membranes.

Of particular interest was the failure to dem-onstrate IgG and betaic-globulin in the glomeruliof four patients in group I whose renal pathologicchanges were moderate to severe. Their clinicaland laboratory features did not distinguish themfrom the other patients with the idiopathic ne-phrotic syndrome; however, they proved resistantto steroid therapy. This absence of immune glob-ulin deposition despite significant renal pathologysuggests that immune globulin deposition is notsimply a secondary phenomenon that ocurs whenglomerular damage is severe. Conversely, as men-tioned earlier, glomerular immune globulin depo-sition has been observed early in the course of dis-seminated lupus erythematosus in patients whoseroutine light microscopic examinations revealedminimal or no abnormalities (19). These ob-servations suggest that the presence or absence ofimmune globulin deposition, as detected by the im-munofluorescent technique, is not directly relatedto the severity of the pathologic changes.

Churg and associates (13) recently separatedpatients with the idiopathic nephrotic syndromeinto two groups on the basis of light and electronmicroscopic findings. One exhibited minimal orno changes by light microscopy and fusion of epi-thelial foot processes by electron microscopy, andthe other showed diffuse membranous thickeningby light microscopy and numerous dense depositsalong the basement membrane by electron micros-copy. These distinctive findings were seen in theinitial studies of the patients, and progression fromlesions of the first type to those of the second was

628

NEPHROTICSYNDROMEOF CHILDHOOD

not observed (13, 19), suggesting different patho-genetic mechanisms. Most of our patients thatshowed no membrane deposition of IgG and betaL-globulin would be classified morphologically withChurg's group that had isolated foot process le-sions. The single child in our study who had dif-fuse membranous glomerulonephritis exhibitedmarked deposition of both IgG and betalc-globu-lin on the glomerular membrane.

None of the patients in groups I and II had de-pressed serum complement levels. Several au-thors have reported such depression in patientswith the nephrotic syndrome (24-26). By con-trast, Ellis and Walton (27) found normal valuesin some of their patients with "uncomplicatednephrotic syndrome," presumably the idiopathicnephrotic syndrome. West, Northway, and Davis(17) used an immunodiffusion technique to meas-ure serum betalo-globulin levels and found normalvalues in patients with idiopathic nephroticsyndrome.

It seems unlikely from these data that immuno-logic mechanisms of the type thus far associatedwith the pathogenesis of renal disease are operativein the most common form of childhood nephrosis.Perhaps immunologic mechanisms that requiregamma globulin and complement in molecularquantities not detectable by these techniques mayproduce the nephrotic syndrome; however, ourinterpretation of the observations herein reportedleads us to favor a biochemical or metabolic de-fect that leads to episodic and potentially reversiblechanges in the permeability of tht glomerular base-ment membrane as the basic abnormality in thenephrotic syndrome of childhood.

SummaryRenal tissue from 35 children with the nephrotic

syndrome was studied by the immunofluorescenttechnique for IgG and beta,(-globulin deposition.Findings were correlated with the clinical courseand with pathologic changes seen by light micros-copy.

Twenty-two children showed no deposition ofeither IgG or betalc-globulin. Most of these chil-dren had minimal renal changes by light micros-copy and responded to steroid therapy. Clinically,the patients of this group corresponded to the mostcommonly seen form of the nephrotic syndromeof childhood.

Seven patients showed a focal type of glomeru-lar IgG and betalo-globulin deposition. Althoughthe exact localization of these deposits was notclear, they were not along capillary basement mem-branes. It was not possible to separate these pa-tients clinically and pathologically from the groupshowing no immune deposits.

A final group of six nephrotic patients whoseglomerular basement membranes showed markeddeposition of IgG and betalcrglobulin was defined.The renal pathologic changes in these childrenwere marked, and none of them responded tosteroid therapy.

Levels of serum complement activity were de-termined in 10 of the patients. Normal valueswere obtained in the patients of the first twogroups.

Weconcluded that immunologic mechanisms ofthe type thus far associated with the pathogenesisof renal disease are probably not operative in themajority of children with the idiopathic nephroticsyndrome.

References1. Drummond, K. N., A. F. Michael, Jr., R. A. Good,

and R. L. Vernier. Immunopathologic studies inthe nephrotic syndrome (abstract). J. Pediat1964, 65, 1114.

2. Michael, A. F., K. N. Drummond, R. L. Vernier, andR. A. Good. Immunologic basis of renal disease.Pediat. Clin. N. Amer. 1964,11, 685.

3. Mellors, R. C., and L. G. Ortega. Analytical pathol-ogy. III. New observations on the pathogenesisof glomerulonephritis, lipid nephrosis, periarteritisnodosa, and secondary amyloidosis in man. Amer.J. Path. 1956, 32, 455.

4. Mellors, R. C., L. G. Ortega, and H. R. Holman.Role of gamma globulins in pathogenesis of renallesions in systemic lupus erythematosus and chronicmembranous glomerulonephritis, with an observa-tion on the lupus erythematosus cell reaction. J.exp. Med. 1957, 106, 191.

5. Burkholder, P. M. Complement fixation in diseasedtissues. I. Fixation of guinea pig complement insections of kidney with membranous glomerulo-nephritis and rats injected with anti-rat kidneyserum. J. exp. Med. 1961, 114, 605.

6. Lange, K., G. Treser, E. Wasserman, and I. Sagel.Routine immunohistology as an aid in the diagno-sis and prognosis of renal diseases (abstract).Ann. intern. Med. 1964, 60, 738.

7. Freedman, P., J. H. Peters, and R. M. Kark.Localization of gamma-globulin in the diseasedkidney. Arch. intern. Med. 1960, 105, 524.

629

DRUMMOND,MICHAEL, GOOD, AND VERNIER

8. Lange, K., M. Wachstein, E. Wasserman, F. Alptekin,and L. B. Slobody. Nephrotic syndrome in thenewborn infant: an immune reaction? (abstract).Amer. J. Dis. Child. 1962, 104, 496.

9. Kouvalainen, K Immunological features in the con-genital nephrotic syndrome. A clinical and ex-perimental study. Ann. Paediat. Fenn. 1963, 9,(suppl. 22).

10. Muller-Eberhard, H. J., U. Nilsson, and T. Aronsson.Isolation and characterization of two B1-glyco-proteins of human serum. J. exp. Med. 1960, 111,201.

11. Michael, A. F., K. N. Drummond, R. A. Good, andR. L. Vernier. Acute poststreptococcal glomeru-lonephritis: immune deposit disease. J. clin. In-vest. 1966, 45, 237.

12. Kabat, E. A., and M. M. Mayer. Complement andcomplement fixation in Experimental Immunochem-istry, 2nd ed. Springfield, Ill., Charles C Thomas,p. 149.

13. Churg, J., E. Grishman, M. H. Goldstein, S. L.Yunis, and J. G. Porush. Idiopathic nephroticsyndrome in adults: a study and classification basedon renal biopsies. New Engl. J. Med. 1965, 272,165.

14. Michael, A. F., K. N. Drummond, R. L. Vernier, andR. A. Good. Immunosuppressive therapy of dif-fuse glomerular disease. In preparation.

15. Wilson, S. G. J., and W. Heymann. Acute glomeru-lonephritis with the nephrotic syndrome. Pedi-atrics 1959, 23, 874.

16. Seegal, B. C., G. A. Andres, K. C. Hsu, and J. B.Zabriskie. Studies on the pathogenesis of acuteand progressive glomerulonephritis in man by im-munofluorescein and immunoferritin techniques.Fed. Proc. 1965, 24, 100.

17. West, C. D., J. D. Northway, and N. C. Davis.Serum levels of betaic globulin, a complement com-

ponent, in the nephritides, lipoid nephrosis, andother conditions. J. clin. Invest. 1964, 43, 1507.

18. Vernier, R. L., H. G. Worthen, and R. A. Good.The pathology of the nephrotic syndrome. J.Pediat. 1961, 58, 620.

19. Drummond, K. N., A. F. Michael, R. A. Good, andR. L. Vernier. Unpublished observations.

20. Hammer, D. K., and F. J. Dixon. Experimentalglomerulonephritis. II. Immunologic events inthe pathogenesis of nephrotoxic serum nephritisin the rat. J. exp. Med. 1963, 117, 1019.

21. Dixon, F. J., J. D. Feldman, and J. J. Vasquez. Ex-perimental glomerulonephritis. The pathogenesisof a laboratory model resembling the spectrum ofhuman glomerulonephritis. J. exp. Med. 1961, 113,899.

22. Heymann, W. Further studies in experimental auto-immune renal disease of rats in Renal Metabolismand Epidemiology of Some Renal Diseases. Pro-ceedings of the 15th Annual Conference on the Kid-ney, J. Metcoff, Ed. New York, National KidneyFoundation, 1964, pp. 217-21.

23. Okuda, R., M. H. Kaplan, F. E. Cuppage, and W.Heymann. Deposition of autologous gamma glob-ulin in kidneys of rats with nephrotic renal dis-ease of various etiologies. J. Lab. clin. Med. 1965,66, 204.

24. Lange, K., L. Slobody, and R. Strang. Prolongedintermittent ACTH and cortisone therapy in thenephrotic syndrome; immunolgic basis and results.Pediatrics 1955, 15, 156.

25. Wedgwood, R. J. P., and C. A. Janeway. Serumcomplement in children with "collagen diseases."Pediatrics 1953, 11, 569.

26. Fischel, E. E., and D. C. Gajdusek. Serum comple-ment in acute glomerulonephritis and other renaldiseases. Amer. J. Med. 1952, 12, 190.

27. Ellis, H. A., and K. W. Walton. Variations in se-rum complement in the nephrotic syndrome andother forms of renal disease. Immunology 1958, 1,234.

630