Segmental glomerulonephritis

Transcription

1 M. S. DUNNILL AND P. R. MILLARD From the Gibson Laboratories, Radcliffe Infirmary, Oxford J. clin. Path., 1975, 28, SYNOPSIS The renal biopsy findings in 40 patients with segmental glomerulonephritis are reported. The term is used to describe a condition in which one or more segments of the glomerular tuft is involved by disease when other segments appear unaffected on light microscopy. The word 'focal' is not used as it may be taken to imply that the changes affect some glomeruli but not others and the evidence for this is not convincing. Segmental glomerulonephritis was a relatively common finding in cases of proteinuria with or without the nephrotic syndrome. The severity of the glomerular changes did not correlate with the ultimate prognosis. On the other hand the tubular and interstitial changes, as assessed by a grading procedure and by point counting, were significantly less severe in those patients who showed clinical recovery than in those who did not. One of the most frequent diagnoses made on renal biopsy material in this laboratory is segmental glomerulonephritis. This term is used to refer to a condition in which a portion of the glomerular tuft is abnormal while the rest of the tuft has a normal appearance. Some writers use the term focal glomerulonephritis to indicate that some glomeruli are normal whereas others are abnormal. This latter opinion, while no doubt true in some instances, represents an essentially two-dimensional view of histology. The average glomerulus is 250,u in diameter and an examination of a 4, section represents a very small sample of such a structure. Often, when serial sections are taken, a glomerulus at one level appears normal but is found to have an abnormal area further down the block. The term focal, or segmental, glomerulonephritis is primarily a histological diagnosis, and, as Heptinstall and Joekes (1959, 1961) showed, the lesion is present in several disease entities. Prominent among these conditions are subacute bacterial endocarditis, periarteritis nodosa, systemic lupus erythematosus, Henoch- Schonlein purpura, Goodpasture's syndrome, Alport's syndrome, focal sclerosing glomerulonephritis, the recurrent haematuria syndrome of childhood, and it is also found in patients without these diseases but with the nephrotic syndrome alone. The present investigation was carried out to ascertain if any indication of prognosis could be given from examination of renal biopsy material. This study Received for publication 11 November was stimulated by the work of Risdon, Sloper, and de Wardener (1968) who demonstrated that in 'persistent glomerulonephritis' there was a significant correlation between renal function and tubular damage. They employed a system of grading in assessing both glomerular and tubular change. In the present investigation a grading procedure was also used but in addition an attempt was made at a more objective assessment of the changes in the biopsy material by the use of point counting (Dunnill, 1968). Material and Methods PATIENTS A total of 43 renal biopsies was seen between the years 1968 and 1972 in which the diagnosis of segmental glomerulonephritis was made. In three cases no satisfactory follow-up study was possible. In the remaining 40 patients a follow-up study was possible and their fate was known for between two and five years after the initial biopsy. The presenting clinical features are given in table I. Patients in whom a specific clinical and pathological diagnosis had been made before biopsy, such as systemic lupus erythematosus, were excluded from this study. As can be seen from table I the majority of patients suffered from proteinuria with or without the nephrotic syndrome. The patients were placed into two groups, A and B. Group A comprised those who were asymptomatic, not maintained on therapy, or 167

2 168 M. S. Dunnill and P. R. Millard 168 M. S. Dunnill and P. R. Millard A Patients recovering from renal symptoms Nephrotic syndrome ('idiopathic') Henoch-Schonlein purpura Proteinuria and hypertension Symptomless proteinuria Persistent proteinuria following acute nephritis Proteinuria following pregnancy Age range 17 to 69 yr (mean 41-7 yr) Male 12 patients: Female 13 patients B Patients who died, are on dialysis, or whose renalfunction has deteriorated Nephrotic syndrome ('idiopathic') Nephrotic syndrome following poststreptococcal nephritis Nephrotic syndrome following pregnancy Henoch-Schfnlein purpura Proteinuria associated with hypertension Proteinuria and chronic renal failure Age range 10 to 72 yr (mean 40 1 yr) Male 8 patients: Female 7 patients Table I Clinical features of cases with segmental glomerulonephritis who had no abnormality other than possibly a 'trace' of protein. Blood urea, total serum proteins, and serum albumin were all within normal limits. Group B contained those who had either remained as when they were first assessed, had received continuous therapy, had deteriorated, or had died. RENAL BIOPSY MATERIAL All biopsies were placed in 4% formaldehyde in saline, dehydrated in alcohol, cleared in xylol, and embedded in paraffin wax. In a few cases a portion of the biopsy was fixed in 4% glutaraldehyde (ph 7T3) at 4 C and subsequently processed for electron microscopy. The paraffin-embedded material was sectioned at 4,u. and stained with haematoxylin and eosin by the periodic-acid-schiff and trichrome methods. Thin (0 5,u) sections were prepared by the method of Burns (1970) and stained by the PAS and silver methenamine methods. ASSESSMENT OF RENAL BIOPSIES Each biopsy was assessed by two observers. The glomeruli, tubules, and interstitial tissue as well as the blood vessels were examined and scored according to a semiquantitative technique, each feature being graded 0 to The features scored in each biopsy are given in table II and approximately correspond to those used by Pollak, Pirani, and Schwartz (1964). A mean score for the glomeruli was arrived at by counting each + as 2 points, adding together all the points recorded by each observer and dividing by 2. An identical procedure was adopted for the tubules and interstitium and for the arteries. There was very little variation between the two observers. POINT COUNTING The point counting method (Dunnill, 1968) employ- Glomeruli Nuclear crowding Necrosis Karyorrhexis Fibrinoid 'Wire loops' Haematoxyphil bodies Hyaline thrombi Fibrosis Tuft adhesions Tubules Casts Atrophy Interstitium Fibrosis Lymphocytic infiltration Arterial vessels Fibrinoid Intimal fibrosis Hyalinization Table II Features assessed in grading renal biopsies ing the Zeiss I integrating eyepiece was used to determine the relative proportions occupied by the glomeruli, tubules, and interstitial tissue on each section. Results CLINICAL FEATURES There was no significant age or sex difference between the two groups ~~ GLOMERULAR CHANGES Nuclear crowdinig This (fig 1) was the most constant finding. Almost invariably it affected the peripheral portion of the glomerular lobule, was associated with obliteration of the lumen of the capillaries in the involved zone, and was accompanied by a fairly striking increase in

3 * 0..' ''_ ^ -r. Fig 1 Glomerulus with localized areas of nuclear crowding. A 24-yr-old man with the nephrotic syndrome. Haematoxylin and eosin x 400. PAS-positive material. This change is often referred to as a focal 'proliferative' change but in our experience mitoses are seldom if ever seen. Frequently there were two or more areas of nuclear crowding in the same glomerulus but intervening normal areas were present and in this series the process never involved the glomerulus in a diffuse manner. Necrosis Necrosis of a segment of the lobule (fig 2) was most frequently seen in Henoch-Schonlein purpura. However, it was not entirely confined to this disease as it was also noted in two cases of an 'idiopathic' nephrotic syndrome which was not associated with systemic lupus. In the areas of necrosis karyorrhexis and fibrinoid change frequently coexisted. 'Wire loops' 'Wire loops' (fig 3) or localized areas of basement membrane thickening were present in eight cases, none of which were clinically diagnosed as suffering Fig 2 Glomerulus with areas of necrosis in lobules. Female of 40 yr with the final diagnosis of Henoch- Schonlein purpura. Haematoxylin and eosin x from systemic lupus erythematosus in that they did not show the LE cell phenomenon or have a positive antinuclear factor in their serum. Haematoxyphil bodies These bodies (fig 3) were seen in six instances and coincided with the wire looping in three cases. Many descriptions of these bodies are vague and unhelpful and published illustrations are not always easy to interpret. In the present investigations they were considered to be distinct from pyknotic or karyorrhectic nuclei and have been taken to correspond to bodies identical with those described by Smith (1955). In haematoxylin-and-eosin sections they appeared somewhat vague in outline, approximately the same size as a red cell, variable in shape though often elliptical, and purplish in colour. Capillary thrombi Capillary thrombi were distinctly uncommon and were seen in isolated glomeruli in only five biopsies.

4 170 M. S. Dunnill and P. R. Millard Fig 3 Fig 4 Fig 3 Glomerulusfrom a man of4j yr with Henoch Schonlein purpura showing (1) localized capillary basement membrane thickening or wire loops, (2) haematoxyphil bodies, one of which is adjacent to a nucleus and arrowed, and (3) karyorrhectic fragments. Haematoxylin and eosin x 480. Fig 4 Portion of a glomerulus with fibrosis of a segment and adhesion between the segment and Bowman's capsule. Haematoxylin and eosin x 400. Focal fibrosis Focal fibrosis of the glomerular tuft (fig 4) and adhesions between the capsule and the tuft were among the most frequent findings being almost universally present though not necessarily prominent in any individual case. Tuft adhesions were found in one or more glomeruli in 36 of the present series of cases. Complete hyalinization of a lobule was seen in cases of some duration. Axial or stalk thickening This (fig 5) was one of the distinctive features described by Hutt and White (1964) in resolving poststreptococcal acute diffuse proliferative glomerulonephritis in African children. However, it is also a feature seen in some of the present cases of focal glomerulonephritis. This is well illustrated in one of our patients with Henoch-Schonlein purpura who was biopsied originally in 1969 and had a repeat biopsy in The 1969 biopsy showed focal necrosis in the glomerular tufts. The biopsy in 1974 was notable for exhibiting a fairly striking degree of axial stalk prominence with an increase in PASpositive material. Tubular changes In some biopsies the tubules were of normal morphology but in others there were variable numbers of eosinophilic casts present usually involving only an occasional tubule (fig 6). Tubular atrophy took the form of either small, deformed, distorted structures lying in a fibrous stroma with an indistinct lumen or,

5 _,_r.- " _..R,.- w 4-_ w wv - - _ *. -w -, Fig 5 Fig 6 Fig 5 Axial or stalk thickening in a glomerulus. Periodic-acid Schiff x 320. Fig 6 Renal biopsy from a woman aged 50 yr who suffered from proteinuria and ultimately developed renalfailure. Nuclear crowding can be seen in the glomerulus. The tubules are showing varying degrees of atrophy and one is dilated due to the presence of a cast. There is interstitial fibrosis and a little lymphorytic infiltration. Haematoxylin and eosin x 135. alternatively, the tubules were dilated, filled with eosinophilic material and lined by a flattened epithelium. In both instances it was not possible to state definitely whether the tubules were originally of the proximal or distal type. INTERSTITIAL CHANGES These consisted of either fibrosis alone or fibrosis associated with focal lymphocytic infiltration. In no case were there aggregates of neutrophil polymorphonuclear leucocytes. The degree of fibrosis was variable and often focal in nature. Usually but not always it was associated with one of the forms of tubular atrophy. VASCULAR LESIONS The main change observed was intimal fibrosis in the interlobular arteries. QUANTITATIVE CHANGES The results are summarized in table III. 171 SCORING METHOD Glomeruli The mean score for the glomeruli in group A was 8-8 and in group B it was 9-5. This difference is not significant (t = 0-52; p > 01). Both groups contained examples of glomeruli exhibiting quite profound areas of segmental necrosis in the tufts and both contained glomeruli which showed only the changes of nuclear crowding. Examination of the score of each of the nine individual characteristics used to make up the total glomerular score, such as nuclear crowding, necrosis, karyorrhexis, etc, did not show that any one characteristic gave a significant indication of prognosis.

6 172 Group A Group B Mean glomerular score 8-8 (sd 4-3) 9 5 (sd 4-1) Mean tubular and interstitial score 4-4 (sd 3 3) 7-6 (sd 5 0) Mean percentage area of interstitium assessed by point counting 26-3 (sd 6 4) 32-9 (sd 12-7) Table III Quantitative assessment of renal biopsies Tubular and interstitial score These two elements were assessed together. The mean score obtained for the patients who recovered was 4-4 and for those who deteriorated was 7-6. This difference is statistically significant (t = 2-29; 0 05 > p > 0 02). POINT COUNTING This provides an objective means of assessing the percentage area of each section occupied by fibrous tissue and tubules. The mean value of the fibrous connective tissue component in the patients who recovered was 26 3 %' and in those who deteriorated the mean value was 32-9 %. This difference is statistically significant (t = 206; 005 > p > 002). The figure given for the interstitial tissue in normal necropsy kidneys by Dunnill and Halley (1973) is 12-8% (sd 5 1). Discussion l ERMINOLOGY The semantics of this condition represent one of the most confusing areas in nephrology. The term 'focal glomerulonephritis' has been used to describe both a clinical disease entity, otherwise referred to as the 'recurrent haematuria syndrome' (Ayoub and Vernier, 1965; Glasgow, Moncrieff, and White, 1970), and a histological lesion in the glomerulus. If the view is accepted that it is probable that every glomerulus is involved to some degree and that the normal appearance of some glomeruli in a single histological section is merely due to the chance of sectioning through a normal plane in a glomerulus which is affected elsewhere, the word 'segmental' (Tighe, 1974) provides a better description. The position is, however, complicated because even this is taking a very limited histological view of the pathological situation. There is some evidence that although histological change may be limited to one portion of the glomerular tuft in fact all of the glomerulus may be affected by the disease process. Berger (1969), in an immunofluorescent study of a very extensive series of renal biopsy material, has demonstrated deposition of IgA, IgG, and 3ic globulin in the mesangial tissues of the glomerulus in areas which appeared normal to light microscopy. M. S. Dunnill and P. R. Millard Davies, Tighe, Jones, and Brown (1973) have also drawn attention to the constancy of diffuse staining for IgA in focal glomerulonephritis associated with the recurrent haematuria syndrome contrasting with the variability of the light microscope findings which showed only occasional segmental glomerular tuft change. Similar conclusions were reported by Evans, Williams, Peters, Sissons, Bolton-Jones, Ogg, Cameron, and Hoffbrand (1973) who found glomerular deposition of C3 and properdin with IgA in cases of idiopathic focal glomerulonephritis and in the Henoch-Schonlein syndrome. We have also seen this phenomenon in biopsies not included in this series. Both light and electron microscope findings have been focal in nature but the immunofluorescent appearances have shown a generalized granular deposition of immunoglobulins, usually IgG, and Pic globulin along the basement membrane. However, the segmental nature of the ultrastructural changes has been particularly striking. Subepithelial electron-dense deposits can be seen associated with fusion of epithelial footplate processes whereas immediately adjacent capillary loops appear quite normal apart from some podocyte deformity (figs 7 and 8). In such circumstances one is forced to the conclusion that either generalized granular distribution of IgG and complement is not itself an indication of disease or that light and electron microscope evidence of structural damage is a relatively late phenomenon. RELATION OF STRUCTURAL CHANGES SEEN ON BIOPSY TO PROGNOSIS There are two indications for percutaneous renal biopsy: (1) to establish a diagnosis, and (2) to indicate, if possible, a prognosis. The emphasis in most reports has been on the glomerular changes. This is because they are often of a dramatic nature and because in patients with the nephrotic syndrome the proteinuria is associated with an excessive leak of protein in the glomerular filtrate. Apart from the work of Risdon et al (1968), tubular and interstitial changes have been largely ignored. This is an anomalous situation as much of the work of the kidney is in the active cellular function of tubular reabsorption whereas glomerular filtration is a relatively passive procedure, to a large extent dependent on a pressure gradient across the glomerular capillary wall.

7 vmt, 0 ;:rsl! :.1. :X Fig 7 Electron micrograph of a glomerular capillary wall, from a man of 65 yr with the nephrotic syndrome, showing electron-dense subepithelial deposits and fusion offootplate processes. x GLOMERULAR CHANGES In the present study quite dramatic changes were observed in the glomeruli in some patients but these provided no indication of the ultimate prognosis. Careful grading of the glomerular lesions showed that in segmental glomerulonephritis there is no consistent relationship between the severity of the changes and the possibility of complete recovery. Patients with the most severe segmental lesions involving necrosis of glomerular lobules sometimes did undergo clinical resolution with disappearance of proteinuria and return to normal renal function. It must be emphasized here that this conclusion is only true for this one histological entity of segmental glomerulonephritis and does not necessarily apply to other forms of histologically identifiable glomerular disease such as diffuse glomerulonephritis or glomerulonephritis associated with epithelial crescent formation. A further point of interest is that changes such as localized thickening of the glomerular capillary 173 basement membrane, karyorrhexis, and even haematoxyphil bodies were seen, either in isolation or together, in glomeruli in cases without any clinical evidence of systemic lupus erythematosus. This we have taken to indicate that the glomerulus has only a certain number of ways of responding to insults and that the disease specificity of any one lesion is strictly limited. Segmental glomerular hyalinization is not in itself an indication of previous segmental glomerulonephritis. It may be found in arteriosclerotic renal disease as illustrated in the serial biopsy studies of five patients reported by Heptinstall and Joekes (1961). The term 'proliferative' is often applied to segmental glomerulonephritis but there is remarkably little evidence to support its use. Nuclear crowding there certainly is in many instances but strict evidence of proliferative activity is limited. Mitoses are extremely rare. Moreover in investigations where nuclear counts have been performed on glomeruli (Bodian, Black, Kobayashi, Lake, and Shuler, 1965) 41

8 174 w... M. S. Dunnill and P. R. Millard Fig 8 Capillary from same glomerulus as in fig 6 with no electron-dense deposits andfairly normal epithelial foot processes. x there has been no evidence of any increase in number of cells and our own unpublished observations confirm this. The cell type is usually considered to be mesangial or sometimes endothelial (Lannigan and Insley, 1965) but epithelial cells are not involved. VASCULAR CHANGES In any renal biopsy the sample is so small that negative vascular findings are of little value with regard to prognosis because of the focal nature of much vascular change. Even positive findings are likely to be of limited significance unless they consist of definite diagnostic features such as fibrinoid necrosis or arteritis. It is interesting that in this series the one case in which a severe degree of vascular change was present in an interlobular artery with both intimal fibrosis and medial hypertrophy follow-up studies showed a good recovery. TUBULAR AND INTERSTITIAL CHANGES The grading procedure demonstrated a small but significant difference in the combined tubular and interstitial changes between those patients who recovered and those who underwent clinical deterioration. It was hoped that the objective method of point counting would demonstrate an even more significant difference. That it did not do so is due to a combination of factors. Tubular damage may take two forms: dilatation with marked flattening of the epithelium and filling of the lumen with eosinophilic material or contraction of the tubular lumen and atrophy of the lining cells. In the former instance the point-counting method will reveal a greatly increased area of the section occupied by the tubules whereas in the latter the tubular area will be markedly decreased. In many patients both processes are present and will tend to cancel one another out. The area occupied by the interstitial tissue appears to be a more reliable guide as to the ultimate fate of renal function in these patients. There was a smaller area occupied by this element in the patients who i

9 recovered than in those who deteriorated. This difference was statistically significant but was not perhaps as great as might have been desired for diagnostic and prognostic purposes. Essentially this is due to the fact that in light microscope studies it is difficult to distinguish between oedema and fibrosis in the interstitium even when using trichrome stains. Previous use of the point count method in the investigation of serial biopsy material from cases of systemic lupus erythematosus has shown that severe interstitial changes can resolve, and that when a relatively large area is occupied by this component in initial biopsy material, subsequent biopsies may show it to have been considerably reduced (Snaith, Holt, Oliver, Dunnill, Halley, and Stephenson, 1973). A further possible explanation for this result is that the ample obtained by needle biopsy of the kidney is not representative of the kidney as a whole but this is not supported by necropsy studies on normal kidneys or kidneys which are the seat of diffuse glomerulonephritis (Dunnill and Halley, 1973). It is indeed a valid criticism in conditions such as pyelonephritis and vascular disease of the kidney. The more likely explanation for the reduction in the interstitial component is resolution of oedema. In segmental glomerulonephritis some of the interstitial change is due to oedema and this no doubt accounts for the fact that point counting does not bring out the difference between the clinical groups more clearly. In spite of the above reservations we do consider that in assessing biopsy material from cases with segmental (focal) glomerulonephritis more attention should be paid to the tubular and interstitial changes. In so doing both a semiquantitative subjective grading of these changes and an assessment of the degree of interstitial fibrosis by point counting can be of considerable assistance in predicting ultimate prognosis. We are grateful to Drs J. G. G. Ledingham and D. 0. Oliver of Oxford, also to Drs B. J. Prout, P. H. Denton, D. H. G. MacQuaide, G. F. Baines, A. Batty Shaw, J. D. H. Slater, J. S. Pryor, Wing Commander C. R. Newman, and Dr Fermie, who have sent renal biopsy material to us and assisted in the follow-up of cases. We are also grateful for the technical assistance in the Gibson Laboratories. References Ayoub, E. M., and Vernier, R. L. (1965). Benign recurrent hematuria. Amer. J. Dis. Child., 109, Berger,J. (1969). IgA glomerular deposits in renal disease. Transplant. Proc., 1, Bodian, M., Black, J. A., Kobayashi, N., Lale, B. D., and Shuler, S. E. (1965). Recurrent haematuria in childhood. Quart. J. Med., 34, Burns, J. (1970). Preparation of thin epoxy resin sections from thick sections of paraffin-embedded material. J. clin. Path., 23, Davies, D. R., Tighe, J. R., Jones, N. F., and Brown, G. W. (1973). Recurrent haematuria and mesangial IgA deposition. J. clin. Path., 26, Dunnill, M. S. (1968). Quantitative methods in histology. In Recent Advances in Clinical Pathology, Series V., edited by S. C. Dyke, pp Churchill, London. Dunnill, M. S., and Halley, W. (1973). Some observations on the quantitative anatomy of the kidney. J. Path., 110, Evans, D. J., Williams, D. G., Peters, D. K., Sissons, J. G. P., Boulton. Jones, J. M., Ogg, C. S., Cameron, J. S., and Hoffbrand, B. I (1973). Glomerular deposition of properdin in Henoch- Schonlein syndrome and idiopathic focal nephritis. Brit. med. J., 3, Glasgow, E. F., Moncrieff, M. W., and White, R. H. R. ( 1970). Symptomless haematuria in childhood. Brit. med. J., 2, Heptinstall, R. H., and Joekes, A. M. (1959). Focal glomerulonephritis: a study based on renal biopsies. Quart. J. Med., 28, Heptinstall, R. H., and Joekes, A. M. (1961). Focal glomerulonephritis. In Ciba Foundation Symposium on Renal Biopsy, edited by G. E. W. Wolstenholme and M. P. Cameron, pp Churchill, London. Hutt, M. S. R., and White, R. H. R. (1964). A clinico-pathological study of acute glomerulonephritis in East African children. Arch. Dis. Childh., 39, Lannigan, R., and Insley, J. (1965). Light and electron microscope appearances in renal biopsy material from cases of recurrent haematuria in children. J. clin. Path., 18, Pollak, V. E., Pirani, C. L., and Schwartz, F. D. (1964). The natural history of the renal manifestations of systemic lupus erythematosus. J. Lab. clin. Med., 63, Risdon, R. A., Sloper, J. C., and de Wardener, H. E. (1968). Relationship between renal function and histological changes found in renal biopsy specimens from patients with persistent glomerular nephritis. Lancet, 2, Smith, J. F. (1955). The kidney in lupus erythematosus. J. Path. Bact., 70, Snaith, M. L., Holt, J. M., Oliver, D. O., Dunnill, M. S., Halley, W., and Stephenson, A. C. (1973). Treatment of patients with systemic lupus erythematosus including nephritis with chlorambucil. Brit. med. J., 2, Tighe, J. R. (1974). Personal communication. 175