April 1979 The Journal o f P E D I A T R I C S

559

Idiopathic rapidly progressive glomerulonephritis with C3 nephritic factor and hypocomplementemia A 7-year-old boy with mild renal failure and signs and symptoms of acute poststreptococcal glomerulonephritis including severe hypocomplementemia had, by renal biopsy, numerous crescents but no deposits in the glomerular capillary loops. Instead, deposits identical in location and composition to those described for children with idiopathic rapidly progressive glomerulonephritis were present. The severe hypocomplementemia was found to be due to high levels of C3 nephritic factor; neither nephritic factor nor hypocomplementemia has been reported in rapidly progressive glomerulonephritis of the idiopathic type. Following prompt therapy with methylprednisolone intravenously, serologic abnormalities disappeared and renal function greatly improved, but a later biopsy showed 50% of the glomeruli obliterated by scarring. The case is of importance not only in indicating that severe hypocomplementemia does not rule out idiopathic rapidly progressive glomerulonephritis but also in adding to the list of diseases in which nephritic factor can be found.

Chades A. Davis, M.D., A. James McAdams, M.D., Robert J. W y a t t , M.D., Judith Forristai, Paul T. McEnery, M.D., and Clark D. West, M.D.,* Cincinnati, Ohio

AT ONSET, idiopathic rapidly progressive glomerulonephritis may closely resemble acute poststreptococcal glomerulonephritis. To distinguish between the two, the serum complement level is to some extent helpful; low levels rule out the idiopathic type of rapidly progressive glomerulonephritis. 1"~ Making the correct diagnosis has become of great importance since recent observations ~'~ give evidence that end-stage disease may be avoided by prompt treatment with either anticoagulants or high doses of methylprednisolone. The present report concerns a boy who was found to be severely hypocomplementemic after sudden onset of glomerulonephritis, identified by renal biopsy as IRPGN. CASE R E P O R T A 7-year-old white boy simultaneously developed pharyngitis, gross hematuria, headaches, and mild periorbital edema in October, 1976. He received a 10-day course of penicillin and "sulfa," but after two weeks of hematuria he was admitted to a local hospital. From the Children's Hospital Research Foundation. *Reprint address: Children's Hospital Research Foundation, Elland Ave. and Bethesda, Cincinnati, OH 45229.

0022-3476/79/400559+05500.50/0 9 1979 The C. V. Mosby Co.

His subsequent course is shown in Fig. 1. He was not hypertensive. The urine contained gross blood, red cell casts, and 3+ protein. The hemoglobin measured 10.7 gm/dl; ESR 38 ram/hour; BUN 13 mg/dl; serum creatinine i.2 mg/dl; C3 16 mg/dl (normal >90 mg/dl); ASO titer 125 Todd units; and Streptozyme titer 100 units. An excretory urogram revealed

Abbreviations used: AGN: acute poststreptococcal glomerulonephritis ASO: antistreptolysin O BUN: blood urea nitrogen C: complement C3NeF: C3 nephritic factor EGTA: ethyleneglycol-bis-(fl-amino-ethyl ether) N,N' tetraacetic acid ESR: erythrocyte sedimentation rate IRPGN: idiopathic rapidly progressive glomerulonephritis

bilaterally enlarged kidneys. By the fourth hospital day, the serum creatinine level had risen to 1.7 mg/dl and he had developed a low-grade fever. The patient was transferred to Children's Hospital where laboratory studies showed a BUN level of 44 mg/dl; serum creatinine 1.7 mg/dl; total serum protein 4.9 gm/dl; and serum

Vol. 94, No. 4, pp. 559-563

560

Davis et al.

The Journal of Pediatrics April 1979

Po/ien/ B 0401 Aqe 7yrs.

)00

Renal_~ - ~ B i ~

~

m0,~

7 - - - R e. n. .~. .".~. . . . . . . .

0

6O Cony. :to

c, f %

0

II)))))lll)lHII)1),Ill ~176 occo,, Foo~ I 9

I

1

)

,,,o

I

,

1,

o

cr,,,,,,o,o,

s~ F

,,0,,, 9 PREDNISONE INDIVIDEDDOSEDAILY

'~ r/

z

Methtlprednisolone I g m iv

]1111

"'

'

'

O

J I

~ PR~'DNISONEON ALTERNATE DAYS

-- --

2

:5

I 4

I 6

I 8

I I0

I ;3

MONTHSAFTERONSET Fig. l. Clinical course of Patient B0401. The horizontal dashed lines indicate the limit of the normal range. albumin 3.0 gra/dl. The C3 level and ASO and streptozyme titers were unchanged. The serum levels of IgG, IgA, and IgM were normal for age. The creatinine clearance was 27.3 ml/minute/ 1.73 m 2 and the 24-hour protein excretion, 1.12 gm. A pea'cutaneous renal biopsy was performed on the third hospital day and the first of five 1.0 grn doses of methylprexlnisolone was given intravenously. Thereafter, signs of acute glomerulonephritis slowly resolved (Fig. 1). Ten months after onset, urine protein excretion was 0.58 gm/24 hours and the creafinine clearance was 81 ml/minute/1.73 m ~. At 16 months, there was mild proteinuria and hematuria, but the BUN and serum creatinine levels were in the normal range. He was never hypertensive. LABORATORY

FINDINGS

Serum eomolement levels. As noted in Fig. 1, serum C3 levels were markedly depressed initially but during therapy with corticosteroids they began to rise; one year after onset, the level remained slightly below the normal range. Measurements of Clq, C4, C5, C6, properdin, factor B, and C3b inactivator in 20 serum specimens obtained over the one-year period since onset have consistently shown levels within the normal range. Nine of the specimens were obtained in the first month of illness. The levels of the above components in serum from the parents were well within the normal range. C3 nephritic factor. Serum was assayed for C3 convert-

ing activity by measurement of the B antigen of C3 ~ before and after incubation of a mixture containing equal volumes of normal human serum and patient serum for 20 minutes at 37~ As shown in Fig. 1, the serum C3 converting activity, expressed as the percentage fall in B antigen concentration, diminished rapidly concurrent with corticosteroid therapy. To identify the serum C3 converting activity with the C3 nephritic factor,' 0.2 ml of serum was layered on a 10 to 40% linear sucrose gradient and centrifuged for 15 hours at 4~ in a Beckman L-2 ultracentrifuge using an SW 50.1 rotor at 36,000 rpm. The 30 fractions, collected dropwise, were quantitated for IgG, IgM, and albumin by radial immunodiffusion. C3 converting activity was determined by adding to 30/~1 aliquots of each fraction, 5 #1 of 0.02 MgCI~, and 15/4 of normal human serum containing 0.095M EGTA from which Clq had been removed by the method of Volanakis and Stroud? B antigen concentration was determined as above. As shown in Fig. 2, the C3 converting activity was found in a broad peak correspondLag to the 6.8S IgG peak. Immune complexes. Circulating immune complexes were measured by a modified solid-phase Clq assay~ in serum samples which had been stored at -70~ The method differed from that originally described in that 0.04M EDTA in the reaction mixture '~ substituted for

Volume 94 Number 4

preheating the serum to 56~ and Tween-20 replaced gelatin as a means of blocking absorption of proteins other than Clq to the polystyrene tube. 11 Results were expressed in terms of the heat-aggregated IgG content of a solution of human IgG which was added to normal human serum. The method is capable of detecting 3 #g/ml of aggregated IgG. The level of immune complexes in 97% of serum specimens from normal subjects, expressed as aggregated IgG, was less than 8/~g/ml. As shown in Fig. 1, the serum level of immune complexes was elevated on the first day of hospitalization, rose rapidly thereafter, but felt to within the normal range in parallel with the fall in C3NeF and concurrent with therapy with corticosteroid. Light microscopy. The initial renal biopsy contained 24 glomeruli, 11 of which had epithelial crescents which were mostly large or circumscribing. In addition, there were numerous focal tuft necroses. Neutrophils in glomerular capillary lumens were common, but there was no mesangial proliferation. There were scattered periglomerular, peritubular, and intratubular infiltrates of neutrophils. On repeat biopsy, eight months after onset, 34 of 68 glomeruli were globally sclerotic and 16 had adhesions to the Bowman capsule or fibrous crescents. There was moderate tubular atrophy associated with interstitial fibrosis and infiltration of small round cells. Ultrastructure. In the initial biopsy, large subepithelial deposits were frequent in the four glomeruli examined and, with rare exception, were located over the nonfiltering portion of the basement membrane (capillary waist) (Fig. 3). Deposit was also demonstrable within folds of the basement membrane in this location (waist basement membrane-related deposit). Breaks of the glomerular capillary wall, associated with crescent formation, were always located at the capillary waist. There were occasional mesangial deposits and minimal evidence of mesangial cellular or matrix proliferation. In the recovery biopsy, the five glomeruli examined by electron microscopy were all focally sclerotic. Persistent subepithelial waist deposits were rare, two being observed. Waist basement membrane related deposit, on the other hand, was frequent. Mesangial deposit was largely replaced by matrix proliferation. Immunofluorescence. Large mesangial-related deposits containing C3, C5, and properdin, corresponding to the subepithelial waist deposits seen in the ultrastructure, were present in the first biopsy. There was, in addition, fragmentary linear fluorescence of the Bowman capsule with anti-C3 and C5. Crescents labeled with antiserum to fibrin. Immunofluorescence with labeled antibody to Clq, IgG, IgA, IgM, and IgE was absent. The recovery biopsy showed by immunofluorescence

Idiopathic glomerulonephritis

561

100

80

,~

70 G(?

IgM

l~G

(19S)

(6flS)

i

1

I

iAL8 (4.3S)

il t

5O

5

I0

15

20

25

50

FRACTION NUMBER

Fig. 2. Sucrose density gradient analysis for C3NeF. The graphs indicate C3 conversion by ultracentrifuge fractions of serum from a normal subject (o .... o) and from Patient B0401 (o---o). The peaks of IgM, IgG, and albumin concentrations are indicated by alTOWS,

no discrete deposits in nonsclerotic glomeruli. In scarred glomeruli, there was an irregular coarse linear label of loops and mesangium with antibody to C3 and a nondescript peripheral label with antibody to IgM. DISCUSSION The history of pharyngitis, gross hematuria, and progressive renal impairment, together with severe hypocomplementemia, suggested that this patient had AGN or the onset of membranoproliferative glomerulonephritis. Renal biopsies in AGN reveal endocapillary glomerular proliferation with deposits on capillary loops containing C3 and often IgG, which correspond to the subepithelial "humps," a hallmark of the acute disease by electron microscopy. 'z In contrast, renal biopsy in the present patient revealed only extracapiUary proliferation. In intact glomeruli there was no evidence of endocapiUary proliferation and, by electron microscopy, subepithelial deposits were limited to the capillary waist; deposits on the loops were very rare. There were no subendothelial or intramembranous deposits. Deposits containing C3, C5, and properdin, presumably those located at the capillary waist, made a mesangial pattern by immunofluorescence. In AGN, the immunofluorescence pattern is that of capillary loop deposits. These same glomerular alterations with deposits of identical composition and position have been previously reported in children with IRPGN. 3

562

Davis et al.

The Journal of Pediatrics April 1979

Fig. 3. Silver-impregnated electron micrograph of a glomeruins from the initial biopsy. Three large subepithelial deposits are present. They are located at the capillary waist (see text) and project into the urinary space. Part of the cytoplasm of a polymorphonuclear leukocyte extends beneath the lower right deposit and is in contact with the underlying basement membrane. (• 5,900.) The second biopsy, eight months after onset, provided further support for the diagnosis of IRPGN. Of 68 giomeruli seen, 75% had a cicatricial change, with 50% globally sclerotic; by immunofluorescence uninvolved g,lomeruli showed no deposits. The observations on this patient are of more than theoretical interest. The serum C3 level is considered helpful in distinguishing patients with rapidly progressive giomerulonephritis of poststreptococcal origin from those whose disease is idiopathic. A normal C3 level is compatible with either diagnosis, but severe hypocomplementemia is usually considered indicative of the more benign, poststreptococcal disease. 1"~The experience with the present patient gives evidence that a low serum C3 level cannot be completely relied upon as a feature distinguishing these two diseases; a definitive diagnosis can be made only by renal biopsy. Early definitive diagnosis of IRPGN is of importance because of evidence that the progress of the disease can be arrested by therapy with anticoagulants ~ or with large doses of methylprednisolone given intravenously.3.,

The C3 converting material in the serum of the present patient met the criteria for C3NeF in that it required Mg § but not Ca +~ to convert C3 and sedimented as a 6.8S protein. 1~ i, Although circulating C3NeF almost invariably produces hypocomplementemia, there has been no direct evidence that it is nephritogenic. TM C3NeF is found most frequently and in greatest abundance in MPGN Type II 1'. 16but may also be found in patients with MPGN Type I,l~ partial lipodystrophy without giomerulonephritis, 1' lupus nephritis, TM and AGN. 19 It has not previously been reported to be present in IRPGN. A puz.zling feature of the illness of the present patient was the persistently reduced concentration of C3 at a level slightly below the normal range, long after C3NeF and immune complexes were no longer detectable in the circulation. There is no explanation for this; measuremerits of complement in the serum of the parents gave no grounds for postulating an inherited partial deficiency of C3. The significance of circulating immune complexes in the serum of this patient during the acute phase of his

Volume 94 Number 4

illness is not clear since experience with this measurement is as yet meager. Although the method detects complexes containing IgG which combine with Clq, neither of these proteins could be detected in the giomerular deposits. It is possible that the immune complexes were altered in the process of being deposited within the glomeruli so that IgG and Clq were no longer detectable. Alternatively, the glomerular deposits may have been derived from circulating complexes undetectable by the method employed, or they may have been formed in situ by an unknown mechanism. Indeed, the complexes in the serum may have resulted from material released from the focally necrotic glomeruli and, as such, been a consequence rather than a cause of the disease. In any event, the circulating complexes had, in all likelihood, no relation to the C3NeF; complexes have sedimentation coefficients greater than 7S ~~but sucrose density gradient analysis of the serum of the present patient revealed C3 converting activity only in a broad 6.8S sedimentary peak.

Idiopathic glomerulonephritis

8.

9.

10.

I I. 12.

13.

14.

REFERENCES

1. Bacani RA, Velasquez F, Kanter A, Pirani C, and Pollak VE: Rapidly progressive (non-streptococcal) glomerulonephritis, Ann Intern Med 69:463, 1968. 2. Anand SK, Trygstad CW, Sharma HM, and Northway JD: ExtracapiUary glomerulonephritis in children, Pediatrics 56:434, 1975. 3. DavisCA, McEnery PT, Maby S, McAdams AJ, and West CD: Observations on the evolution of idiopathic rapidly progressive glomerulonephritis, Clin Nephrol 9:91, 1978. 4. Cole BR, Brocklebank AT, Kienstra R.A, Kissane JM, and Robson AM: "Pulse" methylprednisolone therapy in the treatment of severe glomerulonephritis, J PEDIATR88:307, 1976. 5. Robson AM, Cole BR, Kienstra RA, Kissane JM, Alkaersig N, and Fletcher AP: Severe glomerulonephritis complicated by coagulopathy: Treatment with anticoagulant and immunosuppressive drugs, J PEOXATR90:881, 1977. 6. Ruley EJ, Forristal J, Davis NC, Andres C, and West CD: Hypocomplementemia of membranoproliferative nephritis: Dependence of the nephritic factor reaction on properdin factor B, J Clin Invest 62:896, 1973. 7. Spitzer RE, Vallota EH, Forristal J, Sudora E, Stitzel A,

15.

16.

17.

18.

19.

20.

563

Davis NC, and West CD: Serum C3 lytic system in glomerulonephritis, Science 164:436, 1969. VolanakisJE, and Stroud RM: Rabbit Clq: Purification, functional and structural studies, J Immunol Methods 2:25, 1972. Hay FC, Nineham LJ, and Riott IM: Routine assay for the detection of immune complexes of known immunoglobulin class using solid phase Clq, Clin Exp Immunol 24:396, 1976. Zubler RH, Lange G, Lambert PH, and Miescher PA: Detection of immune complexes in unheated sera by a modified 1~5IClq binding test, J Immunol 116:232, 1976. Svehag SE: A solid phase radioimmunoassay for Clq binding complexes, Scand J Immunol 4:687, 1975. LewyJE, Salinas-Madrigal L, Herdson PB, Pirani CL, and Metcoff J: Clinico-pathological correlations in acute glomerulonephritis, Medicine 50:453, 1971. Vallota EH, Forristal J, Spitzer RE, Davis NC, and West CD: Characteristics of a non-complement dependent C3 reactive complex formed from factors in nephritic and normal serum, J Exp Med 131:1306, 1970. Vallota EH, Spiegelberg HL, Forristal J, West CD, and Mtiller-Eberhard HJ: A serum factor in chronic hypocomplementemic nephritis distinct from immunoglobulins in activating the alternate pathway of complement, J Exp Med 139:1249, 1974. West CD: Pathogenesis and approaches to therapy of membranoproliferative glomerulonephritis, Kidney lnt 9:1, 1976. Ooi YM, Vallota EH, and West CD: Classical complement activation in membranoproliferative glomerulonephritis, Kidney Int 9:46, 1976. SissonsJGP, West RJ, Fallows J, Williams DG, Boucher BJ, Amos N, and Peters DK: The complement abnormalities of lipodystrophy, N Engl J Med 294:461, 1976. Arroyave CM, Wilson MR, and Tan EM: Serum factors activating the alternate complement pathway in autoimmune disease: Description of two different factors in patients with systemic lupus erythematosus, J Immunol 116:821, 1976. Williams GD, Peters DK, Fallows J, Petrie A, Kourilsky O, Morel-Maroger L, and Cameron JS: Studies of serum complement in hypocomplementemic nephritides, Clin Exp Immunol 18:391, 1974. Gabriel A Jr, and AgneUo V: Detection of immune complexes: The use of radioimmunoassays with Clq and monoclonal rheumatoid factor, J Clin Invest 59:990, 1977.