Hypocomplementemic and normocomplementemic persistent (chronic) glomerulonephritis; clinical and pathologic characteristics

Hypocomplementemic and normocomplementemic persistent (chronic) glomerulonephritis; clinical and pathologic characteristics

T h e ] o u r n a l o[ P E D I A T R I C S 1089 Hypocomp lemen tem ic and normocomp lemm tem ic persistent glomerulonepbritis; clinical and pat bo l...

8MB Sizes 0 Downloads 3 Views

T h e ] o u r n a l o[ P E D I A T R I C S

1089

Hypocomp lemen tem ic and normocomp lemm tem ic persistent glomerulonepbritis; clinical and pat bo logic cba racteristics O[ 24 patients with nephritis o[ duration over one year, 17, consistently, over long periods, had normal levels o[ serum fiw-globulin and 7, consistently, had reduced levels. Levels o[ serum complement were normal or reduced in parallel with the file levels. Although no well-defined differences in the clinical course o[ the patients in the two groups were apparent, renal biopsies did show distinctive changes in the [ully developed lesions. In the group with low flw levels, glomerular tu[ts were distinctly lobulated and the thickened capillary walls, on silver staining, were [ound to be nonargyrophilic, giving the appearance o[ poor silver impregnation o[ the basement membrane. In those with normal levels, lobulation was not a notable [eature and basement membranes were well defined by silver impregnation.

Clark D. West, M.D.,* A. James McAdams, M.D., Janice M. McConville, M.B., Ch.B., Nell C. Davis, Ph.D., and Nancy H. Holland, M.D.** CINCINNATI,

OHIO

S T U D I E S I N recent years employing fluorescein-labeled a n t i b o d y " have purFrom the Children's Hospital Research Foundation and the Departments o[ Pediatrics and Pathology, University of Cincinnati College o[ Medicine. Supported by Grants AI-03429 and AI-04472 [rom the National Institute o[ Allergy and ln[ectious Diseases, Public Health Service, and by a grant [rom United Health Foundations, lnc. These observations were previously presented at the Annual Meeting o[ the American Pediatric Society, ]une 1964, Seattle, Washington (]. Pediat. 65: 1022, 1964). ~Address, Children's Hospital Research Foundation, Elland and Bethesda Avenues, Cincinnati, Ohio 45229. **Present address, Department o[ Pediatrics, University o[ Kentucky College o[ Medicine, Lexington, Ky.

ported to show that chronic glomerulonephritis has its origin in an immune reaction. G a m m a globulin has been demonstrated in the glomerular lesions and, more recently, 4' 5 the same technique has indicated the presence of complement components. However, the conclusion that these observations indicate an immune origin of chronic glomerulonephritis is not necessarily correct. In addition to chronic glomerulonephritis, the glomeruli in a number of other types of renal disease of diverse clinical manifestations and origins appear to have a similar affinity for g a m m a globulin. These include lipoid nephrosis, 1 acute nephritis,~, 3-5 periarteritis

1090

December 1965

West et al.

nodosa,~, 3, 4 renal amyloidosis, 1, a systemic lupus erythematosus, 2"~ scleroderma, 3 diabetic nephropathy, a chronic pyelonephritis, 4 Goodpasture's syndrome, ~ nephritis of anaphylactold purpura, 5 hemolytic anemia-uremia syndrome, 5 and malignant nephrosclerosis. 4 Studies employing labeled antibody to Plcglobulin, a complement component, have shown this protein to be present in the lesions not only of acute and lupus nephritis and chronic glomerulonephritis ~, 5 but also in pyelonephritis,4 polyarteritis nodosa, 4 malignant nephrosclerosis,4 Goodpasture's syndrome, 5 and the nephritis of anaphylactoid purpura. 5 The wide variety of lesions in which gamma and Blc-globulin are found dictates caution in ascribing an immune origin to a disease on the basis of labeled antibody observations. Some of the limitations and reservations necessary in interpreting the results of these studies have recently been described. 6 Another approach to the study of the immune origin of renal disease is through measurement of serum complement. A number of workers have shown cornplement to be in low concentration in the serum of patients with acute and lupus nephritis and lipoid nephrosis. 7"~~ In chronic glomerulonephritis, the method has in some hands given confusing results. Ellis and Walton 1~ found the values to vary with time, tending to increase with either clinical improvement or progression of the disease toward the terminal state. Lange, Wasserman, and Slobody,~~ on the other hand, in a study of complement levels on 246 cases of acute nephritis, found 10 cases in which the serum complement remained low and the nephritis persisted with all patients dying of uremia within 4 years. They felt that continuing low levels of complement indicated a poor prognosis. In addition to these patients, there was another group with chronic glomerulonephritis who had signs of renal damage but normal serum complement levels. The prognosis seemed somewhat better in these patients. No data were given on age distribution or renal biopsy observations. Data suggesting two types of chronic nephritis, one with low and

the other with normal complement levels, may also be found in the study by Fischel and Gajdusek. s Measurement of a component of complement in the form of serum Blc-globulin rather than of total complement offers a new means of studying the role of complement in these diseases. Earlier studies have shown this protein to be greatly reduced in concentration in the serum of patients with acute hemorrhagic nephritis 12 and active lupus nephritis, 12"14 and to be in normal concentration in patients with the nephritis of anaphylactoid purpura, lipoid nephrosis, and hereditary hematuria? 2 The results were of interest in cases of chronic nephritis in that, paralleling the observations of Lange and associates, 1~ certain patients had markedly reduced levels while other outwardly similar cases had normal levels? 2 The present paper records in detail observations of these two groups of patients with regard to clinical and laboratory characteristics and changes observed in renal biopsies. Because the biopsies indicated the lesions in these patients to be not only in the chronic stage, but also in a number of patients to be acute, subacute, or indeterminate, the clinical designation of chronic nephritis seems inappropriate and, henceforth, the condition will be referred to as persistent glomerulonephritis. MATERIALS

AND METHODS

Observations were made on 24 patients who had signs and symptoms of nephritis persisting for at least one year. Renal biopsies, obtained on 23 patients, showed in all instances the presence of glomerulonephritis. In the patient not biopsied, there was no doubt about the diagnosis from the clinical course. Patients with the nephritis of anaphylactoid purpura were not knowingly included in the study, although a few patients had a focal nephritis by biopsy suggestive of this disease. Determination of fllc-globulin. The levels of/~lc-globulin were determined either semiquantitatively by inspection of the immuno-

Volume 67 Number 6

Glomerulonephritis

1 09 1

Fig. I. Immunoelectrophoretic pattern illustrating purity of preparation used as a standard in the determination of flla. Normal human serum was placed in the upper well, the flla preparation in a concentration of 0.84 Gin. per cent in the lower well. The antiserum placed in the trough was from goats immunized against whole human serum.

electrophoretic pattern of the serum or, more precisely, by the immunoelectrophoreticprecipitin procedure? 2 In the semiquantitative method, the serum was subjected to electrophoresis in agar by the Scheidegger method and reacted with antiserum from goats immunized either with whole h u m a n serum or with purified fileor fl~a-globulin. Normal control serum was always run on the same slide for comparison. A marked reduction of tim-globulin was indicated by virtual absence of the precipitin arc and lesser grades of reduction by deviation from the normal in length and definition of the arc as well as in its distance from the electrophoretic axis. In most instances subnormal levels of file-globulin are easily recognized by the method. 12 In case of doubt, measurement was made by the immunoelectrophoretic-precipitin method. The modification of the immunoelectrophoretic-precipitin method which allowed its use for measurement of fl~c-globulin was described previously? 2 A change in the consumption of antibody by this protein occurs as it is converted to fllA by aging. Hence, to be consistent, measurements have always been made when the protein was in the ill, form, and the assumption made that the level of fl~A-globulin in aged serum reflects that of tim-globulin in fresh serum. Conversion of tim to fl~A was brought about by allowing the serum to stand either at room temperature for 3 or more days, or a combination of 8 hours or more at room tempera-

ture and a minimum of 1 week at -10 ~ C. The method was standardized using preparations of purified ilia-globulin prepared by one of us (N. C. D.) as described previously. TM Kjeldahl analysis was used to determine the protein content of the preparations. The immunoelectrophoretic pattern indicating the purity of the preparation finally used for the standardization is shown in Fig. 1. Determination of complement. Complement in 50 per cent hemolytic units per milliliter (C'H50) was determined according to the method of Kabat and Mayer, 15 as modified by Wedgwood and Janeway2 A commercially available,* washed and pooled 10 per cent suspension of sheep cells was used, and diluted to a 5 per cent suspension on the day of the test. The hemolysin (amboceptor) was commercially available as a 50 per cent glycerinated sheep red cell antiserum.t It was titrated by the method of Kabat and Mayer. 15 Dilutions of serums varied from 1 : 2.5 for low complement levels to 1:20 for normal levels. Optical densities of the hemoglobin present in the supernatants were read at a wave length of 541 m/, on a Beckman D U spectrophotometer. The optical densities were corrected for the density of the appropriate dilution of the serum being assayed to avoid falsely high values with colored or lipemic serums. The 50 per cent hemolytic unit titer was ~Baltimore Biological Laboratory (70-001Q). tBaltimore Biological Laboratory (75-001J).

West et al.

1 092

December 1965

RESULTS

calculated using the von K r o g h formula and plotting the results on log-log paper. A minim u m of 3 points was used to describe the line for each serum. If 3 points were not determined, or if those determined did not readily describe a straight line, the results were discarded and the serum reassayed, often at a new dilution. Serum specimens were stored in solid CO2 until use. Renal biopsies. Thirty-three renal biopsies were studied. T w o of these were obtained by open operation; the remainder were needle biopsies. T h e tissues were fixed immediately in 10 per cent formalin. Sections were made at 6 m/, for hematoxylin and eosin stain and the periodic acid-Schiff procedure, and at 1 to 2 m/, for the Jones-methenaminesilver procedure? ~;

Clinical and laboratory observations.

Grouping of patients according to serum ~,e levels. I n Fig. 2 serum /?1e-globulin levels are shown for 24 patients, expressed in terms of pin-globulin, plotted against the time after onset of the nephritis. Values indicated by circles were obtained on 7 patients, those indicated by squares, on 17 patients. For all patients, values for specimens obtained during or shortly after therapeutic courses of immunosuppressive agents in the form of cyclophosphamide, mechlorethamine, or 6-mercaptopurine, have been omitted and, in the group with low levels, values during prednisone therapy are indicated by open circles. A good separation is apparent between the two groups. T h e separation is

9 - I ? patients with normal 81c levels 0

7 patients with low 8 1 c Levels w i t h prednisone therapy

9 -

without prednisone therapy

240 220-

- 220 &&

200

-

-

9

9

~

180 -

9m <

160m

140

1

mm

9

m: l"

-

120-

w

I00 -

(z~

80-

EE

9

'1

9

9

9 I

- 180

- 160

9140

,, n

200

9

~,L

9120

J,

-I00 -80

Norrnol

Subiec~s

60-

0 o

0

0 0

40-

0 9

9

o

eO0 0

0 o

go I

n

J

0

o

O0

20-

oo

~

I

Ol~ 9

9

i

i

.

O0

9 ,

f

i

. . . . . .

3

2 Time

0

0

in y e a r s

,~

9

. . . . . . .

5

;

. . . . . .

7

since onset of n e p h r i t i s

Fig. 2. Serum fllA concentrations in 24 patients with persistent nephritis. Seventeen of the patients, indicated by squares, had normal serum levels and 7, indicated by circles, had reduced levels. The range for normal subjects is shown on the right. Values are plotted against years since apparent onset of nephritis.

Volume 67 Number 6

Glomerulonephritis

1 093

trophoretic pattern. Inclusion of these semiquantitative values in graph of Fig. 1 would not have altered the distribution of the points. Since all patients with persistent glomerulonephritis encountered in a 7 year period are included, it is apparent that the incidence of patients with normal levels is more than twice that of patients with reduced levels. Age at onset and sex distribution were about the same in the two groups. Boys constituted 71 per cent of the group with low levels and 59 per cent of the group with normal levels. In 19 of the 24 patients the disease was first noted in the winter season from November to April. It is of interest that 7 of the patients with normal file levels had their onset in the period from January to March, 1963, inclusive.

particularly striking when only the values obtained in the absence of prednisone therapy are considered. Subnormal values have not been observed at any time in any of the 17 patients with normal levels, and normal levels have never been seen in the absence of therapy in any of the group with reduced levels, even though some patients have had the disease for as long as 7 years. The age at onset, duration of the disease, time of renal biopsy, and frequency of estimation of file-globulin levels are shown in Fig. 3 for the 17 patients with normal fl~eglobulin levels. The same data for the 7 patients with reduced levels are shown in Fig. 4. Determination of file-globulin was made at least twice on each patient. Values for some specimens were determined semiquantitatively by estimation from the immunoelec-

B

Age

at o n s e t . . .

5

B

ell~

V.J.

B

B =Renal

Biopsy

9

inlilti

I

II

.L = T i m e

of

~31c G l o b u l i n

determination

ill I t I ]I

12

D= Death

B

14

III

r

B

12

Inlillill

16

I

II]ll

~

15

li

o

M.W

c~

D.C.

o

B.H.

I

II B

B , IIIII B I

6

B

I

B

M.L.

B

B

I

T.K.

5v o

R.H.

o

C.H.

o

WH.

o

B

5

.I

I

I]1

B 12

IIIIII II

]

I

V.C.

B 14

,

~f

IIII

o 12 Jilt

I

I I I Li

A.B I I

12

II

II

L.S.

d J o

M.M. ~ B

t

L I t

J.T.

-~

B

5 12 9

9

S

, II

I

I 8 I

I B

II I

I

R.R

II ]

R.M. I

N.P. ~

YEARS

Fig. 3. Data for 17 patients with normal file levels, including year of onset, age at onset, and times of measurement of file-globulin and of renal biopsies.

West et al.

1 094

December 1965

B= Renal Biopsy -L = Time of Bic Globulin determination

Age at onset .,. 9

B

II

III P.D.

B

J2 IIIIIIIIIIIIIIlllllll B 7' 5

B

7.

B

B [11

IIIII

I I I

B

I

I!1

II

9

II

I

'

I

'

'58

I

I

'59

J

I

I

I[ I

IIINll i l i

I

'60

'63

~

s.s.

o~

R,W.

a'

G.C.

d

L,B,

I

'62

'61

R,S.

R.L.W. d

I tl

B

B i

B

1957

I

B INIIIIIlll I I I liN I I 1

I

'64

YEARS

Fig. 4. Data for 7 patients with low serum/3m levels.

Comparison o[ serum levels o[ iliA-globulin with hemolytic complement. Levels of hemolytic complement and of fltA were determined on the same serum specimens from 23 normal subjects, 6 patients with acute nephritis, 6 patients with persistent nephritis and normal fl~c levels, and 6 patients with persistent nephritis and low fl,c levels. The normal range for hemolytic complement, as determined in the normal subjects, ranged from 13 to 38 C'Hs(, units per milliliter. The results in all subjects are plotted in Fig. 5. Whereas the points scattered considerably, deviating from a straight-line relationship, there was general agreement among values in the various groups. All patients with acute or persistent nephritis with low fl,, (fl,c) levels had subnormal levels of hemolytic complement, whereas patients with persistent nephritis and normal fl~a (fl~c) levels were well within tile normal range for complement. The relationship indicates that the terms hypocomplementemic and normocomplementemic glomerulonephritis would be appropriate in referring to patients with low and normal /?w levels, respectively. Comparison o[ clinical course and laboratory observations in patients with normo- and hypocomplementemic persistent nephritis. In

view of the distinct differences in complement and file levels in these two groups of patients, it seemed of interest to examine the clinical and laboratory data to identify other features which might distinguish the two groups. At the onset, the majority of patients in both groups were thought to have acute glomerulonephritis. In the normocomplementemic group, 10 patients were initially diagnosed as having acute glomerulonephritis in severe to moderately severe form, while in 7, the initial diagnoses were nephrotic syndrome, subacute nephritis, pyelonephritis, and urinary tract bleeding. In the hypocomplementemic group, the initial event was diagnosed as acute nephritis in 6 of the 7 patients. One was thought to have urinary tract bleeding. In general, the onset was more typical of acute nephritis in the hypocomplementemic patients, and suspicion of the presence of a disease more serious than juvenile acute nephritis was aroused late in the course. In fact, in 3 of the hypocomplementemic patients, the acute nephritis was thought to subside completely until symptoms returned several months later concomitant with an acute infection. In contrast, in the normo-

Volume 67 Number 6

Glomerulonephritis

1 095

3009 Normal subjects o Persistent nephritis normal ('JlC 9 Acute nephritis Z~ Persistent nephritis

250-

now Oic

200"

9

j

I a~

150"

E <

i

Lower limit Normal Range

100.

50-

O

JO

A

9

oo

9 9

0

9

9 O

I

9 O

9

9

i~

9

9

9

J i I~_ Lower limit

9 /%

Normal Range 0

0

I

I

5

I0

I

i

I

I

15 20 25 30 35 Hemolytic Complement CIHsoUnits/ml

I"

4O

Fig. 5. Comparison of simultaneous C'Hs0 and fl,A (fl,c) values in normal subjects and in patients with nephritis. complementemic group, the course often suggested much earlier the presence of a disease more serious than acute nephritis; the blood urea nitrogen remained elevated, the patients remained clinically ill, and in 11, the edema present at the onset became more severe, constituting a prominent part of the clinical picture. The majority of patients had an infection preceding the onset of their nephritis. In 12 of the normocomplementemic group, a respiratory infection occurred 2 to 3 weeks before the onset and was typically a sore throat, fever, and cough, often with "flulike" symptoms. T w o other patients had similar infections but nephritis developed during the infection. In 3 there was no known preceding infection. In the hypocomplementemic group, 4 had a history of a preceding respiratory infection and one had onset during an infection. T w o had no known antecedent infections. The data available on titers of antistreptolysin O at the time of onset of the nephritis

for 5 of the 7 hypocomplementemic patients and for 10 of the 17 normocomplementemic patients are given in Table I. In the hypocomplementemic group, one patient had definite evidence of preceding streptococcal infection with a rise in titer to over 500 units. Another had a moderate elevation of 250 units, and 3 had relatively low titers. In the normocomplementemic group, high titers were seen in 4 patients, and in 2 of these a significant rise in titer could be demonstrated. T w o others had values of 250 units, but the remaining were in the range of 166 units or below. It is felt that the data are not sufficiently complete to allow conclusions as to the incidence of streptococcal infection prior to the onset of the nephritis. It should be emphasized, in considering the events at the onset, that the first clinical manifestation of the nephritis may not represent the actual onset of the disease. In one hypocomplementemic (G. C.) and one normocomplementemic patient (J. T . ) , it was found that hematuria and proteinuria had

1 096

December 1965

W e s t et al.

Table I. ASO titers at apparent onset of hypo- and normocomplementemic persistent nephritis Time alter Time alter ASO titer onset onset (Todd nephritis injection (da~,s) (days) Patient units) Hypocomplementemic persistent nephritis G.C. 125 2* No infection 100 5* No infection S.S. 125 23* 23 R. L.W. 5O 2 8 R.W. 25O 2 9 R.S. 166 l 22 500 21 42 Normocomplementemie persistent nephritis N.D. 250 11 18 333 19 26 L.S. 500 37 51 V.C. 250 2 23 A.B. 166 3 No infection J.T. 250 5* 5 833 41" 41 B.H. 125 9 27 V.J. 166 14 31 50 29 46 W.H. 833 3 5 R.P. 166 6 No infection M.L. 500 5 No infection 833 13 No infection *Actual onset of nephritis may have antedated

clinical

onset,

been present on routine urinalyses performed 2 and 4 years, respectively, before the apparent onset of the disease. In another hypocomplementemic patient (S. S.), pallor and unusual fatigue had been noted for a year before the apparent onset. It is of interest that in 2 of these patients, the first manifestation of disease occurred concurrent with, rather than following an infection, and in the other (G. C.), there was no preceding or concurrent infection. As the disease developed, the clinical manifestations were not sufficiently distinctive to give a clear-cut differentiation between the two groups. As mentioned above, edema was more often prominent early in the normocomplementemic group, but moderate to severe edema also developed at some time in the course of the disease in 3 of the 7 hypocomplementemic patients, and was present at the onset in one patient. Hypertension

was somewhat more frequent in the hypocomplementemic group and was particularly apt to occur in association with steroid therapy. Of interest was the fact that one patient in the hypocomplementemic group (R. L. W.) had his signs of nephritis gradually abate without change in file-globulin levels. Proteinuria and hematuria slowly disappeared after having been present for 16 months, and renal function, as measured by urea clearance, returned to normal, yet serum fllA levels remained markedly depressed at levels of less than 12 mg. per cent. The boy had had no therapy for his disease. In another hypocomplementemic patient (P. D.), the only evidence of nephritis consisted of 3 episodes of gross hematuria and occasional microscopic hematuria occurring over a period of 19 months. There was only minimal proteinuria. Renal biopsy 13 months after onset showed acute nephritis. In none of the patients was there evidence of lupus. Preparations to detect lupus erythematosus cells were consistently negative, and antinuclear antibody was absent from the serum in all. Except for the nephritis, clinical evidence of lupus was lacking. Evidence for this disease was particularly sought in the hypocomplementemic group because lupus nephritis is also accompanied by low complement and file levels. The patients in the two groups differed in only one laboratory parameter and this difference was not invariably present. Certain of the patients with hypocomplementemia tended to have anemia out of proportion to the degree of their azotemia. In 4 of the 7 in this group, levels of hemoglobin below 10 Gm. per cent were often observed at times when the blood urea nitrogen was not greater than 25 mg. per cent. Red cell indices in these patients indicated the cells to be normochromic and normocytic. There was no evidence of hemolysis; reticulocyte counts in 2 of the patients gave values of less than 3 per cent, and haptoglobin was invariably present on the immunoelectrophoretie pattern. A bone marrow examination on the patient with the most severe anemia showed

Volume 67 Number 6

a mild erythroid hyperplasia, and assay of his serum for erythropoietin activity gave normal results, e Histopathologic changes observed in renal biopsies. It was found that detailed listing of observable alterations in the 33 biopsies obtained on 23 of the patients was of little value in showing points of similarity and difference in the two groups. Of greater value was a classification as to the stage of the nephritis, and certain key observations which distinguished the histopathologic changes in the two groups. Classification of stage of nephritis. The stage of disease has been determined from the appearance of the hematoxylin and eosin stained preparations according to the following criteria. Acute glomerulonephritis indicated an endothelial proliferative lesion of the glomerular tuft, usually with some degree of polymorphonuclear leukocytic exudation. In several instances, the glomerular tuft showed unusual lobulation but the appearance did not deviate sufficiently from that of typical acute glomerulonephritis to justify another term. Two forms of subacute glomerulonephritis were observed. In both, there was thickening of the capillary walls. In one form adhesions were inconspicuous or absent and there were no crescents, but endothelial proliferation and exudation were of a degree to make any other designation unreasonable. In the other form, designated as subacute glomerulonephritis with crescents, the usual features of adhesions and crescents were present. Chronic glomerulonephritis indicates a lesion of sufficient magnitude to be associated with tubular atrophy. The glomerular lesion was variable, either embodying hyaline thickening of the tuft capillary walls or hyaline and reticular obliteration of the tuft. When thickening of the tuft capillary walls was the principal alteration with little evidence of endothelial proliferation, the lesion

*Erythropoletln assay was performed by the hypertransfused mouse method by Dr. James W. Linman, Veterans Administration Research Hospital, Chicago 11, Ill.

Glomerulonephritis 1 0 9 7

is referred to as chronic membranous glomerulonephritis. In chronic focal glomerulonephritis there was focal involvement of the glomeruli as well as focal intraglomerular lesions. Indeterminate nephritis was used in the literal sense for lesions which were not classifiable as chronic glomerulonephritis but which were not readily assignable to any other category. Generally, the biopsies in such instances showed some degree of endothelial proliferation without exudation and irregular thickening of the tuft capillary wall, frequently with axial thickening.

Histopathologie observations in hypoeomplementemic patients. As noted in Table II, renal biopsies of 3 patients showed acute glomerulonephritis. In one of these (P. D.), the lesion was characterized only by mild endothelial proliferation and slight exudation, but in the others (S. S. and R. L. W.), there were severe endothelial proliferation, prominent exudation, and distinctly lobulated tufts (Figs. 6 and 7). The tuft capillary walls appeared irregularly thickened in Patient S. S. In the Jones-methenaminesilver preparation, both Patients S. S. and R. L. W. showed some irregularity of the silver impregnation of the tuft, with splitting of the capillary basement membrane or production of a pericapillary space (Figs. 8 and 9). In our experience, splitting of the basement membrane rarely occurs in classical acute glomerulonephritis. In a subsequent biopsy, 15 months later, S. S. showed a subacute giomerulonephritis without crescents (Fig. 10). The glomeruli were large, conspicuously lobulated, and showed prominent endothelial proliferation. These features and the frequent occurrence of a hyaline appearance of the more central portion of the lobulated tuft were very reminiscent of what has been set forth as chronic lobular glomerulonephritis in adults. 16 The Jones preparation of Patient S. S. showed more irregular argyrophilia than it had in the acute stage with a more definite pericapillary space seeming to contain a brownish deposit. In the axial portion of the tuft there was an elaboration of argyrophilic fibrils.

West et al.

1 098

December 1965

T a b l e I I . S t a g e of n e p h r i t i s in 23 p a t i e n t s w i t h p e r s i s t e n t g l o m e r u l o n e p h r i t i s

Hypocomplementemic persistent nephritis ~ Patient I First biopsy ISeeond biopsy[ Third biopsy P.O. R.S. S.S. R. L . W . R.W. G.C. L.B.

AGN SGN AGN AGN SGN SGN SGN (cresc)

SGN SGN SGN

SGN

NormocompIementemic persistent nephritis Patient I First biopsy ISecond biopsy V, J. M.L. M.W. D.C. B.H. T. K, R.H. C.H. W.H. V.C. A.B. L.S. J.T. R.P. R.M. N.D.

CFGN Allergic angiitis Indeterminate Indeterminate AGN CGN (memb) Indeterminate Indeterminate SGN CGN Indeterminate SGN (cresc) Indeterminate Indeterminate CGN SGN (cresc)

CFGN CFGN CFGN

CGN (memb) Indeterminate

AGN, acute glomerulonephlitis SGN, subacute glomerulouephritis SGN (cresc), subacute glomerulonephritls with crescents CGN, chronic glomerulonephritis CFGN chronic focal glomeruloneph~itis CGN (mernh), chronic membranous glomerulonephr tis

Fig. 6

Fig. 7

Fig. 6. Needle biopsy of S. S. with hypocomplementemic disease i0 weeks after apparent onset. The changes are those of acute glomerulonephritis with endothelial cell proliferation, mild exudation, and definite lobulation. The capillary walls seem somewhat thickened. (Hematoxylin and eosin.) Fig. 7, Open renal biopsy on R. L. W. with hypocomplementemic disease 10 months after onset of nephritis. Acute glomerulonephritis is present with endothelial cell proliferation and slight exudation. (Hematoxylin and eosin.)

Volume 67 Number 6

Fig. O

Glomerulonephritis

1 099

Fig. 9

Fig. 8. Jones-methenamine-silver preparation of the biopsy on hypocomplementemic Patient S. S. while in the stage of acute nephritis, as shown in Fig. 6. The portion of the glomerulus, here seen under oil immersion, shows apparent splitting of the capillary basement membrane producing a pericapillary space. Fig. 9. Jones-methenamine-silver preparation of the biopsy on hypocomplementemic Patient R. L. W. in the stage of acute nephritis as shown in Fig. 7. Splitting of the basement membrane is present, similar to, but not as severe as seen in S. S., Fig. 8.

Four other patients (G. C., R. W., R. S., and L. B.) showed subacute glomerulonephritis. In one of these, L. B., the initial lesion was characterized by crescent formation and adhesions, but a subsequent biopsy 65 months later showed complete disappearance of adhesions and crescents (Fig. 11). The glomerular tufts were moderately lobulated, the tuft capillary walls were thickened, and there was endothelial proliferation. In the Jones preparation there was irregular to poor argyrophilia with a nonargyrophilic deposit in the capillary wall and an axial proliferation of fine argyrophilic fibrils (Fig. 12). The others, G. C., R. W., and R. S., were characterized by prominently lobulated, hypercellular glomeruli, greatly thickened tuft capillary walls, and, frequently, central hyaline zones (Fig. 13). Argyrophilia was greatly reduced in R. S. with splitting of the basement membrane, apparently by deposition of nonargyrophilic material. The capillary walls

in R. W. and G. C. showed only fragmentary silver impregnation in the Jones preparation (Figs. 14 and 15). In the case of R. W., the 3 biopsies obtained over a period of 44 months showed progressive diminution in the degree of argyrophilia accompanied by increasing deposition of nonargyrophilic material. The periodic acid-Schiff preparations did not add to the above observations and will not be considered. In summary, renal biopsies in 7 patients with hypocomplementemia showed 5 to have a subacute glomerulonephritis, all characterized by increased cellularity and a notable thickening of the capillary walls. A unique alteration in the degree of argyrophilia of the capillary wall was observed, resulting, apparently, from splitting of the basement membrane by a nonargyrophilie material. Silver impregnation of the capillary wall was virtually obliterated by this process in 2 of the patients and greatly reduced in the other

1 1 O0

West et al.

December 1965

Fig. 10. Needle biopsy of S, S. with hypocomplementemic disease 18 months after apparent onset. Subacute glomerulonephritis without crescents is now present. There is marked endothelial cell proliferation, the glomerular tufts are lobulated, and hyaline zones are developing in some of the lobules. Capillary walls are definitely thickened. (Hematoxylin and eosin.)

3. ~ T h e glomerular lesion in 4 of these 5 patients was very similar to that described for chronic lobular glomerulonephritis, especially as manifested by prominent tuft lobulation with a tendency to central hyaline zones. In one patient, this lesion followed one of acute glomerulonephritis. T w o patients, whose renal biopsies were classified as acute glomerulonephritis, have persistent hypocomplementemia but have not been rebiopsied.

~'Experlence subsequent to the preparation of this manu~ script has showll that fixation of needle biopsies f r o m patients with hypocomplementemic nephritis in Bouin'~ solution results in a less distinctive appearance of the glomerular tuft after formalin fixation. T h e silver deposits in the capillary wall with Bouin's are more dense and, because of this, are poorly rc~olved,

Histopathologic observations in normocomplementemic patients. As seen in Table II, the renal lesions in this group of patients included chronic focal glomerulonephritis, chronic membranous glomerulonephritis, and allergic angiitis, as well as lesions in the acute, subacute, and indeterminate stages. Only selected examples from these patients are detailed here. T h e biopsy in one patient, V. C., resembled in the hematoxylin and eosin preparation the fully developed lesion of the low /31c group. T h e glomerular tufts were lobulated and hypercellular, and the capillary walls thickened (Fig. 16). There were, however, conspicuous differences apparent in the Jones-methenamine-silver stain, in that the capillary basement membranes, although severely thickened, remained argyrophilic.

Volume 67 Number 6

Fig. I1

Glomerulonephritis

1! 0 1

Fig. 12

Fig. 11. Needle biopsy of hypocomplementemic Patient L. B., 5~/2 years after onset, showing subacute nephritis without crescents. Note the slight lobulation, the marked endothelial cell proliferation, and the developing hyaline zones in some of the lobules. Capillary walls are thickened. (Hematoxylin and eosin.) Fig. 12. Jones-methenamine-silver preparation of the biopsy of L.B., 5y2 years after onset, shown in Fig. 11. In this oil immersion view of a glomerulus, the capillary walls are seen to be irregularly argyrophilic with a nonargyrophilic deposit present apparently within the capillary wall.

Fig. 13. Needle biopsy of hypocomplementemic Patient R. W. 5 years after onset, showing subacute glomerulonephritis without crescents. Note the marked lobulation with central hyaline zones in the lobules and the marked endothelial cell proliferation. (Hematoxylin and eosin.)

Fig. 14. Oil immersion view of glomerulus from biopsy of R. W. 3 years after onset. Jones-methenamine-silver preparation. Nephritis is in the subacute stage. The greatly thickened capillary walls show virtually no silver impregnation. Only punctale and fine fibrillar staining can be seen.

1102

December 1965

West et al.

Fig. 15. Jones-methenamine-silver preparation of biopsy of hypocomplementenfic Patient G.C. taken 3 years after apparent onset. There is almost complete obliteration of the usual argyrophilia of the capillary wall by a nonargyrophilic material, resulting in marked thickening of the wall.

There was an external, nonargyrophilic deposit on the basement membrane and a severe axial argyrophilic reticular proliferation (Fig. 17). Two other patients had chronic glomerulonephritis of the membranous type. In one, R. M., the glomerular tufts were not unusually lobulated (Fig. 18) but had moderately severe argyrophilic thickening of the capillary basement membranes (Fig. 19). In the other, T. K., the glomeruli were moderately lobulated but showed insignificant endothelial proliferation (Fig. 20). There was a severe argyrophilic thickening of the tuft capillary basement membrane and an external deposition of alternating argyrophilic and nonargyrophilic material producing a beaded appearance characteristic of the advanced lesions of chronic membranous glomerulonephritis (Fig. 21) .17 In only one case did the renal biopsy show the feature of splitting of the tuft capillary

basement membrane seen in the hypocomplementemic group. This was the second biopsy of Patient R. H. It differed from those of the hypocomplementemic patients in that the hyperceilularity was not so great, lobulation was minimal, and the overall picture was that of chronic focal glomerulonephritis. Another patient who had chronic focal glomerulonephritis (B. H.), had the characteristics of acute glomerulonephritis at the time of initial biopsy 14 months earlier (Figs. 22 and 23). In summary, the variety of lesions found at biopsy in the 16 normocomplementemie patients did not suggest any common pattern of evolution. In only a few cases were there any of the constellation of histopathologic features observed in the hypocomplementemic patients, and in these cases, other morphologic features differed significantly. For example, in some biopsies a nonargyrophilic deposit was present, but it was located on the epithelial side of the capillary wall. Furthermore, it did not reduce the argyrophilic properties of the basement membrane as was the case in the hypocompIementemic patients. It can be concluded that by use of routine hematoxylin and eosin staining and of silver impregnation techniques, biopsies in these two groups are distinguishable, particularly in the subacute and chronic stages. DISCUSSION

The hypocomplementemia in renal disease has been ascribed to fixation of complement to complexes containing gamma globulin present in the glomerular lesions. Evidence that the same explanation applies to the reduction of serum file-globulin was discussed previously.12 The complex responsible for this fixation could result from reaction of antibody with a fixed renal antigen or with a circulating antigen. In the latter case, the immune complex would be deposited in glomerular capillary walls by the mechanism responsible for this event in serum sickness. Alternatively, the nephritis might result from deposition in glomeruli of complement fixing aggregated gamma globulin without antibody

Volume 67

Fig. 16

Number 6

Glomerulonephritis

1 10 3

Fig. 17

Fig. 16. Needle biopsy of normocomplementemic Patient V. C. one year after onset. Subacute nephritis is present with moderate lobulation of the tufts, marked endothelial cell proliferation, and thick capillary walls. There are no central hyaline zones. (Hematoxylin and eosin.) Fig. 17. Jones-methenamine-silver preparation of biopsy of normocomplementemic Patient V . C . shown in Fig. 16. Oil immersion view of glomerulus shows thickened capillary basement membranes which are argyrophilic. There is a nonargyrophilic deposit on the epithelial side of the basement membrane interrupted by argyrophilic projections.

Fig. 18

~ig. 19

Fig. 18. Needle biopsy of normocomplementemic Patient R. M. 44 months after onset. Chronic membranous glomerulonephritis is present, with thickened capillary walls and slight endothelial ceil proliferation. (Hematoxylin and eosin.) Fig. 19. Jones-methenamine-silver preparation of biopsy of normocomplementemic Patient R . M . shown in Fig. 18. Oil immersion view of glomerulus shows that the capillary walls are generally thickened and retain their argyrophilia.

1 I 04

West et al.

Fig. 20. Needle biopsy of normocomplementemic Patient T.K. 4 months after onset, showing chronic membranous glomerulonephritis. There is moderate lobulation and marked thickening of capillary wails. (Hematoxylin and eosin.)

activitylS, 19 as postulated by Michael, Drummond, Vernier, and Good. ~ A choice among these possibilities cannot be made from the information available at present. On the other hand, differences in serum complement and serum /Sac-globulin in the two groups of patients herein reported might conceivably be ascribed to a reduced ability to produce file-globulin, perhaps as an inborn defect, independent of the presence of nephritis. This seems unlikely, however, in that in immunoelectrophoretic patterns on over 6,000 patients, lowered levels have not been seen except in renal and hepatic disease. Further, in one patient in the hypocomplementemic group, levels of tim-globulin rose to the normal range in conjunction with immunosuppressive therapy, 2~ indicating normal ability to produce this protein. Evidence against excretion of file-globulin in the urine as the cause of the reduced levels was discussed previously. 12 Recent studies in which actual measurement of file

December 1965

Fig. 21. Jones-methenamine-silver preparation of biopsy of normocomplementemic Patient T.K. shown in Fig. 20. Oil immersion view shows marked thickening by an external deposit of nonargyrophilic material, interspersed by spiny argyrophilic projections from the thickened argyrophilic basement membrane. and fl,A have been made indicate only minute amounts present in the urine. In fact the rate of excretion in normal subjects is of the same order of magnitude. 21 Differences in file-globulin levels and complement might alternatively be ascribed to individual differences among these patients in their reaction to a common injury. T h a t the disease is basically different in the two groups of patients is, however, strongly suggested by the distinctive appearance, morphologically, of the renal lesion in patients with hypocomplementemia. It cannot be stated with certainty that all the hypocomplementemic patients will eventually develop a glomerular lesion characterized by lobulation, hypercellularity, and nonargyrophilic hyalinization of the tuft capillaries, but this is our impression. We have reviewed in consultation 3 other patients, not part of this report, who are hypocomplementemic and

Volume 67 Number 6

Fig. 22

Glomerulonephritis

1 1 05

Fig. 23

Fig. 22. Open biopsy of normocomplementemic Patient B.H. 3 weeks after onset, showing acute glomerulonephritis manifested chiefly by moderately severe endothelial cell proliferation. (Hematoxylin and eosin.) Fig. 23. Jones-methenamine-silver preparation of biopsy of normocomplementemic Patient B. H. shown in Fig. 22. Oil immersion view shows normal capillary wall without evidence of basement membrane splitting.

whose subacute renal lesions are of an entirely similar nature. It is hazardous, if not impossible, to relate the biopsy findings in these examples of hypocomplementemic nephritis to observations by others on the histopathology of chronic nephritis. However, descriptions of chronic lobular glomerulonephritis by Allen, 22 and by Jones aG essentially agree with what appears to be the fully developed lesion in hypocomplementemic nephritis. Churg and Grishman, 17 in speaking of subacute glomerulonephritis of the intercapillary type, indicate the eventual formation of central lobular scars as described in the present series. Steiner and associates, 23 in studies on chronic glomerulonephritis in adults, make special note of a situation in membranous glomerulonephritis (lipoid nephrosis of adults) in which the initial lesion is characterized by exudation of a proteinaceous material between the basement membrane and the epithelium of the capillary loop. The silver preparation in their Fig. 16 depicting this lesion is quite reminiscent of the changes

in our cases of hypocomplementemic nephritis. This light microscopic illustration represents to us "splitting" of the capillary wall basement membrane. Whether chronic lobular glomerulonephritis has the same meaning to all observers is uncertain. One should not assume that such patients, on examination, will show persistently lowered serum file-globulin levels. The present observations, however, as well as those of previous workers strengthen the concept that hypocomplementemic nephritis has etiologic distinctiveness. Granted that fixation or destruction of complement on complexed or aggregated gamma globulin is responsible for the hypocomplementemia, certain puzzling questions arise when various facets of these observations on norrno- and hypocomplementemic subjects are considered. Although no attempt has been made in this or other studies to correlate the serum levels of fl~e-globulin with the amount of this protein deposited in the renal glomeruli as determined by immunofluorescent techniques, it would appear

1 10 6

West et al.

from the data available that such a study might reveal a number of discrepancies. It would be expected that in the normocomplementemic patients, the type most frequently encountered, complement would not be deposited or would be deposited to a limited extent in the lesions; yet, in all cases of subacute or chronic nephritis which have been examined by imn'mnofluorescent techniques for the presence in the glomeruli of file_globulin~, 5 or of complement, 24 results have been definitely positive. Similarly the immunohistologic complement fixation test 2"~ was always positive in such cases. It is obvious that further study will be needed to clarify the apparent discrepancies in these two approaches. A feature of interest is the duration of the hypocomplementemia as measured by the depressed flae levels. In acute nephritis the reduction of file-globulin is transient, '2 apparently because the eornplexed or aggregated gamma globulin is present for a limited period. In contrast, in the hypocomplementemic patients reported here, very low levels of file-globulin were present over periods measured in years. The source and nature of the complex which continues to fix fl,c-globulin so completely that the protein is nearly absent in the serum for long periods can only be speculated on. Continued antigen release would appear necessary to maintain a high rate of antibody formation. Release of antibody into the circulation at a constant rate great enough to produce, on reaction intravascularly with soluble antigen, the amount of complex required to inactivate fi,e-globulin as rapidly as it is formed could account for the disease. However, the nature and source of the antigen available in the quantity necessary to produce "persistent serum sickness" is difficult to visualize. Reaction of antibody with a fixed glomerular antigen seems plausible, but over a long period of time, a source of the antigen necessary for the continuous stimulation of antibody formation must be found. One possible mechanism might be the gradual destruction of renal tissue as a result of the phlogistic effect of the complement-binding complexes.

December 1965

It should be noted, however, that a severe necrotizing lesion would not be necessary for this. In the 2 patients, R. L. W. and P. D., whose low /71c levels have persisted over a long period while they exhibit minimal evidence of renal disease, the renal lesion consists of only a mild acute nephritis. In these patients, the phlogistic effect of complementbinding complexes appears to be minimal, an observation which will bear further investigation. Perhaps most attractive as an explanation of the prolonged hypocomplementemia is the concept that aggregation of gamma globulin is occurringY, 18 Such aggregates would bind complement and, depositing on the glomeruli, could result in an inflammatory reaction as suggested by Michael and associates. 5 Against this concept is the observation that with immunosuppressive therapy, a rapid rise of file-globulin levels to normal has been observed in one patient, as though antibody formation had indeed been suppressed3 ~ Without knowledge of the exact mechanism of the hypocomplementemia, one can nevertheless speculate on the etiology of hypocomplementemic disease. It is most attractive to postulate that the initial event is acute poststreptococcaI nephritis. In favor is the fact that, clinically, the initial disease in a number of patients in this group resembled closely typical juvenile, acute nephritis, and in 3, the renal biopsies had many features in common with acute nephritis. Lange and associates 1~ assumed the original disease in the hypocomplementemic patients they observed to be classical acute nephritis. Why a small minority of patients with acute glomerulonephritis would continue to form complement-binding complexes and remain hypocomplementemic is not known. Severity of the disease is not a factor. In no hypocomplementemic patient was the initial disease considered to be unduly severe, and in several cases, it was very mild. It should be noted that, although acute glomerulonephritis seems to be a logical origin for hy-pocomplementemic disease, it is not possible from the present data to conclude definitely that these cases represent a continuation of

Volume 67 Number 6

this disease. It is possible that the hypocomplementemia and the nephritis developed insidiously without clinical manifestations and came to light only when an intercurrent infection made the nephritis manifest. The patients with normocomplementemic nephritis in the present series probably represent a mixture of a number of renal diseases not yet separable into entities. To claim differing etiologies based on differences in clinical course is hazardous because of the variability of individual response which is possible from a common insult. Suffice it to say that these patients did differ among themselves in clinical manifestations, but the details do not seem appropriate to the present paper. Probably the best evidence that several entities are represented resides in the variety of histopathologic changes in the biopsy specimens. It seems likely that, at a minimum, 3 or 4 etiologically different renal diseases are represented. Their nature can only be speculated on. They would all appear to have lesions not dependent on the presence of a complementbinding complex. Some might arise from an immune reaction in which delayed hypersensitivity is the primary mechanism, or alternatively, the mechanism could be that concerned with homograft rejection. On the other hand, there is no direct evidence from the present study to ascribe the origin to an immune reaction at all. SUMMARY

Of 24 patients with nephritis persisting over one year, 17 were found to have normal levels of serum fllc-globulin and 7 had reduced levels. In a given patient, the levels were consistent over a long period of time. In those with low levels, normal levels were at no time encountered unless immunosuppresive therapy had been used, and in those with normal levels, reduced concentrations were not seen. Serum complement levels, measured as C'H~0, in general paralleled the concentration of serum fllc-globulin. Renal biopsies demonstrated the disease in both groups to be in various stages, from

Glomerulonephritis

1 1 07

acute to chronic, and an acute lesion was seen in one patient more than a year after onset. Hence, the two conditions have been designated as hypocomplementemic and normocomplementemic persistent glomerulonephritis. Renal biopsies showed distinctive changes in the two groups. In the hypocomplementemic patients in the subacute stage, there were thickened capillary walls with moderate to severe lobulation of the tufts with a central hyaline zone as has been described for chronic lobular nephritis by others. Staining by the Jones-methenarnine-silver procedure showed a nonargyrophilic deposit in the capillary walls, often with splitting of the wall. The deposit progressively encroached on and replaced the basement membrane so that the normal argyrophilia of this structure was present only irregularly or, in some cases, was completely absent. Biopsies in the normocomplementemic patients showed lesions in all stages as well as the specific lesions of allergic angiitis, chronic focal glomerulonephritis, and chronic membranous glomerulonephritis. The impression was that a number of disease entities were represented in this group. Rarely were lobulation or splitting of the capillary wall by a proteinaceous deposit, changes distinctive for the hypocomplementemic biopsies, seen in the normocomplementemic group. In cases in which these changes were seen, other features, distinctive for the hypocomplementemic subjects, were absent. Comparison of clinical and laboratory observations on the hypo- and normocomplementemic patients showed no distinct differences. Onset tended to be more typical of mild to moderate acute glomerulonephritis in the hypocomplementemic group and the impression was that the hypocomplementemic disease represented a continuation of acute nephritis, but this could not be proved. Marked edema occurred early in the clinical course in a number of the normocomplementemic subjects. The hypocomplementemic subjects had a greater tendency to hypertension, especially during cortico-

1 1 08

December 1965

W e s t et al.

steroid therapy. A n e m i a was also disproportionately c o m m o n in the hypocomplementemic patients a n d was often out of proportion in severity to the degree of azotemia. None of these differences, however, were sufficiently well defined to allow certain distinction between the two groups based on clinical or laboratory parameters. O n e hypocomplementemic patient has had spontaneous i m p r o v e m e n t of his nephritis with disappearance of all signs a n d symptoms, b u t continues to have severely reduced serum /3m-globulin levels. A n o t h e r has had only slight h e m a t u r i a for over a year a n d moderately reduced p~c-globulin levels with a biopsy showing, after a year, a mild acute nephritis. I n these patients, the absence of i n f l a m m a t o r y reaction contrasts sharply with the extent of c o m p l e m e n t fixation which is a p p a r e n t l y occurring. We are indebted to June Baker, Judith Forristal, Nancy Ulmer, and Evelyu Wright for assistance in determining P,A levels, to Betty Wilson for aid with the photomicrography, and to Eva Lachman, and Betty Lou Wilson for preparation of the manuscript. Our thanks also go to the many physicians in the Cincinnati area and to those in Ohio, Kentucky, and Indiana who directed the patients to us. REFERENCES 1. Mellors, R. C., and Ortega, L. G.: Analytical pathology. III. New observations on the pathogenesis of glomerulonephritis, lipid nephrosis, periarteritis nodosa, and secondary amyloidosis in man, Am. J. Path. 32: 455, 1956. 2. Mellors, R. C., Ortega, L. G., and Holman, H. R.: The role of gamma globulins in pathogenesis of renal lesions in systemic lupus erythematosus and chronic membranous glomerulonephritis, with an observation on the lupus erythematosus cell reaction, J. Exper. Med. 106: 191, 1957. 3. Freedman, P., Peters, J. H., and Kork, R. M.: Localization of gamma globulin in the diseased kidney, Arch. Int. Med. 105: 524, 1960. 4. Lachmann, P. J., Muller-Eberhard, H. J., Kunkel, H. G., and Paronetto, F.: The localization of in vivo bound complement in tissue sections, J. Exper. Med. 115: 63, 1962. 5. Michael, A. F., Drummond, K. N., Vernier, R. L., and Good, R. A.: Immuno!ogic basis of renal disease, Pediat. Clin. North America 11: 685, 1964. 6. Unanue, E., and Dixon, F. J.: Experimental

7.

8.

9. 10.

11.

12.

13.

14.

15. 16. 17. 18.

19. 20.

21. 22.

glomerulonephritis. IV. Participation of complement in nephrotoxic nephritis, J. Exper. Med. 119: 965, 1964. Lange, K., Graig, F., Oberman, J., Slobody, L., Ogur, G., and LoCasto, F.: Changes in serum complement during the course and treatment of glomerulonephritis, Arch. Int. Med. 88: 433, 1951. Fischel, E. E., and Gajdusek, D. C.: Serum complement in acute glomerulonephritis and other renal diseases, Am. J. Med. 12: 190, 1952. Wedgwood, R. J. P., and Janeway, C. A.: Serum complement in children with "collagen diseases," Pediatrics 11: 569, 1953. Lange, K., Wasserman, E., and Slobody, L. B.: The significance of serum complement levels for the diagnosis and prognosis of acute and subacute glomerulonephritis and lupus erythematosus disseminatus, Ann. Int. Med. 5:]: 636, 1960. Ellis, H. A., and Walton, K. N.: Variations in serum complement in the nephrotic syndrome and other forms of renal disease, Immunology 1: 234, 1958. West, C. D., Northway, J. D., and Davis, N. C.: Serum levels of Blo globulin, a complement component, in the nephritides, lipoid nephrosis, and other conditions, J. Clin. Invest. 43: 1507, 1964. Morse, J. H., Muller-Eberhard, H. J., and Kunkel, H. G.: Antinuclear factors and serum complement in systemic lupus erythematosus, Bull. New York Acad. Med. 38: 641, 1962. Seligmann, M.: Immunoelectrophoretic study of the serum during the course of systemic lupus erythematosus, in Grabar, P., and Burtin, P., editors: Immunoelectrophoretic analysis, applications to human biological fluids, Amsterdam, 1964, Elsevier, chap. 16, pp. 209-214. Kabat, E. A., and Mayer, M. M.: Experimental immuno-chemistry, ed. 2, Springfield, Illinois, 1961, Charles C Thomas, Publisher. Jones, D. B.: Nephrotic glomerulonephritis, Am. J. Path. 33: 313, 1957. Churg, J., and Grishman, E.: Subacute glomerulonephritis, Am. J. Path. 35: 25, 1959. Ishizaka, T., Ishizaka, K., and Borsos, T.: Biological activity of aggregated y-globulin. IV. Mechanism of complement fixation, J. Immunol. 87: 433, 1961. Ishizaka, K.: Gamma globulin and molecular mechanisms in hypersensitivity reactions, Progr. Allergy 7: 32, 1963. West, C. D., Holland, N. H., McConville, J. M., and McAdams, A. J.: Immunosuppressire therapy in persistent hypocomplementemic glomeruloneohritis and in lupus nephritis, J. PEmAT. 67: 1113, 1965. West, C. D., and MeConville, J. M.: Unpublished observations. Allen, A. C.: The kidney, medical and surgical disease, New York 1962, Grune & Stratton, Inc., chap. 7, pp. 256-270.

Volume 67 Number 6

Glomerulonephritis

23. Steiner, J. W., Movat, H. Z., Yendt, E. R., and Slater, R. J.: Studies on chronic glomerulonephritis. II. A concept of the morphogenesis of the disease based on light and electron microscopic observations, Canad. M. A. J. 86: 195, 1962. 24. Freedman, P., and Markowitz, A. S.: Gamma globulin and complement in the diseased kidney, J. Clin. Invest. 41: 328, 1962. 25. Burkholder, P. M.: Complement fixation in diseased tissues. I. Fixation of guinea pig complement in sections of kidney from humans with membranous glomerulonephritis and rats injected with anti-rat kidney serum, J. Exper. Med. 114: 605, 1961.

APPENDIX REPRESENTATIVE

CASE

SUMMARIES Hypocomplementemic persistent nephritis. R. L. W., born March 6, 1957, had been retarded in development since the age of about 7 months when he apparently had an episode of encephalitis. At the age of 5 years, in April, 1962, he was admitted to Children's Hospital with the diagnosis of acute glomerulonephritis. For a week prior to admission, he had had a cough and was pulling at one ear, and 2 days preceding admission, he had had rhinitis and fever. Admission was prompted by the observation of rusty-colored urine. On admission there was an acute otitis media but no edema or hypertension. The blood urea nitrogen was 52 rag. per cent; antistreptolysin-O titer, 50 units; hemoglobin level, 10.5 Gin. per cent; and the ilia level, 9 rag. per cent. A chest x-ray revealed pneumonia. In addition to hematuria, there was a trace of proteinuria which was not present on succeeding days. During the 13 days' hospitalization, the blood urea nitrogen dropped to 29 rag. per cent and hematuria diminished to microscopic. On several occasions in the 6 months after discharge, the local physician noted continuing microscopic hematuria. In July, 1963, he was admitted after a series of illnesses including otitis media, stomach ache, and vomiting. Pneumonia was again present. On urinalysis there were 10 to 12 red blood cells per high-power field (uncentrifuged) and a trace of proteinuria. Edema and hypertension were absent. The hemoglobin was 8.7 Gm. per cent and the fl~Alevel was 12 mg. per cent, but blood urea nitrogen, serum cholesterol, and total protein were normal. Two preparations to demonstrate lupus erythematosus cells were negative,

1 1 09

and gastric washings were negative for siderocytes. Nevertheless, biopsies of the lung and kidney were performed since Goodpasture's syndrome was suspected. The lung tissue was normal, but the renal biopsy showed a severe proliferative and moderately exudative lesion of the glomeruli without adhesions or crescents and with only slight basement membrane thickening (Fig. 7). In the Jones preparation, the capillary basement membranes showed areas of splitting and somewhat irregular impregnation by silver (Fig. 9). For 2 months after discharge, until March, 1963, tests performed by the parents at home for urinary protein were positive at trace to 2 plus levels approximately every other morning, and for hemoglobin, about every third morning. Respiratory infection transiently increased the frequency and intensity of both. However, in the third to seventh months after discharge, positive tests for protein and hemoglobin became less frequent, and periods of 3 to 6 weeks would pass during which the urine was normal. Subsequently, abnormalities have been seen even less frequently. In September, 1964, during an episode of appendicitis and pneumonia occurring simultaneously, proteinuria went only to the 1 plus level and hematuria became microscopic, and both disappeared quickly with convalescence. The blood urea nitrogen has remained in the normal range, and urea clearance is 84 per cent of normal. However, of 10 measurements of the serum fllA levels made in the 2 year period during which the nephritis has been quiescent, none have been higher than 10 rag. per cent. L. B. was born Nov. 14, 1947. In early December, 1956, this 9-year-old girl had onset of puffiness of the face and hands, weight gain, and red urine. An upper respiratory infection had been present 2 weeks previously. She was admitted to a local hospital where gross hematuria and a 4 plus proteinuria were noted. On treatment with penicillin, the hematuria subsided but proteinuria persisted and she became more edematous. The blood urea nitrogen at onset was 61 mg. per cent, falling to 34 mg. per cent one week later. Two weeks after onset the cholesterol was found to be 360 mg. per cent and the total protein, 4.35 Gm. per cent. She was then referred to the Children's Hospital. On admission she was obviously edematous and had a blood pressure of 173/104. Within

1 1 10

W e s t et al.

a few days after admission ascitic fluid was noted. Proteinuria was 2 plus to 3 plus and there was a microscopic hematuria. Urine specific gravities did not exceed 1.015. The hemoglobin level was 12.8 Gin. per cent. The blood urea nitrogen was t6 mg. per cent, but cholesterol had risen to 632 rag. per cent, and total protein diminished to 3.7 Gin. per cent. The patient was placed on a low salt diet and given antihypertensives. She slowly improved clinically. A renal biopsy showed moderately severe proliferative changes and slight exudation in the glomeruli with numerous adhesions and crescents. The tuft basement membrane appeared moderately thickened in routine stains; in the Jones preparation there was irregular silver impregnation, splitting of the basement membrane, and deposit of nonargyrophilic material in the pericapillary space. After 20 days of hospitalization during which the patient lost 9 pounds in weight, she was discharged to be readmitted 2 weeks later for therapy with A C T H . Urinalysis again showed a 4 plus proteinuria and microscopic hematuria, but specific gravities and blood urea nitrogen were in the normal range. The hemoglobin was 11.5 Gin. per cent. She was given A C T H in a dose of 55 units per day, and 6 days after the start, she received 0.4 rag. per kilogram of mechlorethamine over a 4 day period. A C T H was then stopped and she was discharged, unimproved. At home she remained at bed rest for 6 months without further treatment. Gradually, evidence of the nephrotic syndrome disappeared. As she became ambulatory, she occasionally had periorbital edema but felt well, and 4 years after onset she was on unlimited activity and lived a normal life. However, proteinuria and hematuria persisted and occasionally, a moderate hypertension was noted. In June, 1962, at the age of 14 years, 51/2 years after onset, she was found to have proteinuria ranging from 1 plus to 4 plus and a microscopic to gross hematuria. Hemoglobin was 14.5 Gm. per cent. The blood urea nitrogen and cholesterol were normal. Total protein was 5.4 Gm. per cent. Creatinine clearance gave normal values. The /?1A-globulin level was 7 mg. per cent. A second renal biopsy showed disappearance of the adhesions and periglomerular fibrosis present at the first biopsy. Thickening of the tuft capillary wall was more severe than before

December 1965

(Fig. 11 ) and a considerable diminution of silver staining of the tuft had occurred due to deposition of nonargyrophilic material in the capillary wall (Fig. 12). The tufts were mildly lobulated. There was a rare sclerotic glomerulus. The patient was discharged on no medication. In the 2 ~ years since that admission she has continued at full activity and is leading a normal life. However, proteinuria is still present at the 3 plus to 4 plus level, but hematuria has been absent. She has been generally normotensive, but, recently, more consistent blood pressure elevations have been noted. Blood chemistries are only mildly altered. The blood urea nitrogen is 18.5 mg. per cent; cholesterol, 330 rag. per cent; and total protein, 5.5 Gm. per cent. /31A-globulin is still 7 mg. per cent. Normocomplementemic persistent nephritis. B. H. was born Jan. 8, 1951. In late February, 1963, this 12-year-old girl had the flu with pharyngitis and fever to 102 ~ F., for which she was given 2 injections of penicillin. In early March, edema, accompanied by proteinuria and later hematuria, was noted. Twelve days after the onset of edema she was admitted to the Children's Hospital. On admission she was obviously edematous but there were few other abnormalities on physical examination. There was no hypertension. Laboratory studies showed the hemoglobin level to be 12 Gm. per cent; serum cholesterol, 310 mg. per cent; total protein, 4.1 Gm. per cent; and the blood urea nitrogen, 36 mg. per cent. Urinalysis revealed a 4 plus proteinuria and up to 10 red blood cells per high-power field (uncentrifuged). A urea clearance gave values of 38 and 41 per cent of normal function. The serum fllx level was 101 mg. per cent and the antistreptolysin-O titer, 125 units. Three preparations to detect lupus erythematosus cells were negative, and antinuclear antibody was not present in the serum. Gastric washings were negative for siderocytes. A renal biopsy showed moderate proliferative and exudative changes in the glomeruli but no thickening of the tuft capillary walls which showed normal silver impregnation of the basement membranes (Figs. 22 and 23). The child was treated with diuretics and with prednisone. The latter was given in a dose of 60 rag. daily and was combined with mechlorethamine after 16 days of therapy. The treatment appeared to have little effect on the proteinuria and the hematuria. With the steroid

Volume 67 Number 6

therapy, the blood urea nitrogen rose to a peak value of 52 mg. per cent and the serum cholesterol to 548 rag. per cent. After discharge she was placed on prednisone on an intermittent schedule in a dose of 80 rag. per day. She did not do well, however. Severe edema developed, necessitating treatment with spironolactone, chlorothiazide, and stringent salt restriction. The serum cholesterol fell to normal within 3 weeks after discharge, but thereafter it began to rise, and by July, 1963, reached a level of 804 mg. per cent. Total serum protein remained in the range of 4.2 to 5.0 Gm. per cent; BlA-globulin, 160 to 180 rag. per cent; and the blood urea nitrogen, 30 to 40 mg. per cent. A moderate hypocalcemia developed. The hematuria diminished to microscopic and eventually ceased altogether, but proteinuria continued unabated. In early July, 1963, a course of 6-mercaptopurine was given. The dose started at 75 mg. per day and was later increased to 100 rag. per day. The course lasted for 18 days and the drug was stopped when the white blood cell count dropped to 3,700 per cubic millimeter. During this time, prednisone was given continuously in a dose of 30 rag. once daily. There was no immediate effect of this therapy, but 2 months later the urea clearance had risen to 50 per cent of normal, and the cholesterol had fallen to 475 rag. per cent. However, edema, easy fatigability, and lassitude continued. In the fall of 1963, the child was put on an intermittent schedule of prednisone administration, and in September and October, a course of cyclophosphamide was given. During and after this course of therapy, she was subjectively improved but laboratory parameters remained the same. However, by June, 1964, the blood urea nitrogen was well in the normal range; serum cholesterol was 286 mg. per cent; and edema was much less of a problem although total protein was still 4.7 Gm. per cent. Prednisone had been continued on an intermittent schedule during this time. A second renal biopsy in June, 1964, showed considerable variation in glomerular involvement. Some glomeruli appeared normal while others showed capsular scaring, tuft scarring, and adhesions. Focal tubular atrophy was present. The diagnosis was chronic focal glomerulonephritis. Since that time, the patient has been off all medication and has led an active life without edema. By February, 1965, all labora-

Glomerulonephritis

111!

tory parameters were in the normal range, but proteinuria was present in the trace to 1 plus range. V. C. was born Oct. 17, 1949. In May, 1962, at the age of 12x/2 years, this girl developed an upper respiratory infection with pharyngitis. Three weeks later her eyes were noted to be puffy, and hematuria and proteinuria developed. She was admitted to a hospital in another city. The blood pressure was initially elevated at 150/ 90 but dropped to normal rather quickly. The antistreptolysin-O titer was 250 units. Hemoglobin level was 13.4 Gm. per cent and the blood urea nitrogen, 14.8 mg. per cent. During this hospitalization she lost 13 pounds in weight. The only therapy was penicillin. In September, 1962, after an uneventful summer, urinalysis showed a 3 plus proteinuria and marked hematuria. Further studies gave approximately the same results as on the admission 4 months previously. In addition, serum cholesterol was found to be 320 rag. per cent and total protein, 5.1 Gin. per cent. A preparation to demonstrate the lupus erythematosus cell phenomenon was negative. In May, 1963, one year after onset, the girl was hospitalized at the Children's Hospital. In the year since onset, her disease had not changed despite markedly curtailed activities. The child appeared to be healthy but had mild pretibial edema. The blood pressure was normal. Proteinuria was 2 plus to 3 plus and hematuria was microscopic. A urine concentration test resulted in a peak value of 1.018, but most morning urine specimens had specific gravities in the range of 1.009 to 1.014. The blood urea nitrogen was 16 mg. per cent; cholesterol, 228 mg. per cent; and total protein, 4.6 Gm. per cent. A urea clearance' gave values 23 per cent and 33 per cent of normal. Two preparations to demonstrate lupus erythematosus cells were negative and the serum was negative for antinuclear antibodies. Serum /?,A-globulin levels ranged from 100 to 140 mg. per cent. A renal biopsy showed moderate proliferative changes in the glomeruli, occasional adhesions, prominent lobulation of the tuft, and moderately severe thickening of Bowman's capsule (Fig. 16). There were no crescents. The tuft capillary walls were moderately thickened and in the Jones preparation showed thickening of the basement membrane with an external deposit of nonargyrophilic material (Fig. 17). There was a severe cicatrization of a stalklike mesangium.

1112

West et al.

Of particular note was the widespread focal tubular atrophy. The diagnosis was chronic glomerulonephritis. The patient was treated with a course of prednisone combined after 2i days with mechlorethamine. The course of therapy was uneventful and not complicated by hypertension. The blood urea nitrogen increased during prednisone therapy to as high as 52 mg. per cent, and serum cholesterol to 350 mg. per cent. She was continued on 30 mg. prednisone once daily for 3 months after discharge. During this time the blood urea nitrogen fell to normal, cholesterol

December 1965

to 310 mg. per cent, and the proteinuria to a trace to 1 plus. Microscopic hematuria was still present. In January, 1964, urea clearance was 80 per cent of normal. Daily prednisone has subsequently been stopped and an intermittent schedule of prednisone instituted with a daily dose of 80 mg. She has returned to normal activity. In May, 1964, all medication was stopped, and in August, serum chemical values were normal, but a trace of proteinuria and a slight microscopic hematuria persisted.