Key words: Glomerulitis, pathology, renal allograft, morphometry
University Institute of Pathology*, Ćarhus Kommunehospital, and Department
of Nephrology, University Hospital in Skejby, Ćarhus, Denmark
Department of Pathology**, and Department of Nephrology, University of Alberta, Edmonton, AB, Canada.
Address reprint requests to: Niels Marcussen, M.D. University Institute of Pathology Ćarhus Kommunehospital 8000 Ćarhus C Denmark Fax +45 89 49 37 00Acknowledgement:
Early transplant glomerulitis (ETG) is characterized by the presence of an increased number of mononuclear inflammatory cells and endothelial cell swelling in glomerular capillaries. The lesion usually occurs in the first months after transplantation. The first description of ETG was given in 1965 by Hamburger et al. . Porter et al. (1968)  mentioned that 78% had subendothelial humps or mesangial deposits usually positive for IgM and C3. Most of these specimens were, however, from periods later than 1 year post transplant. Axelson et al. (1985)  showed, based on the investigation of 576 graft biopsies that 8% had glomerulitis of mild to severe degree. No correlation to CMV infection was found. The lesion was only present in acute rejection but there was no correlation between the grade of rejection and of glomerulitis. They concluded that glomerulitis is a distinct form of acute rejection. We have previously shown that ETG is positively correlated with acute rejection but has not been shown to have had any effect on the graft survival . In our experience the lesion often resolves by itself. It appears to be associated with two very different conditions - cyclosporine nephrotoxicity  and anti-Class I antibody-mediated rejection [4a].
The purpose of the present investigation was to estimate the glomerular volume in biopsies with and without glomerulitis to see whether the lesion leads to an increase in glomerular size. Furthermore, we wanted to characterize the glomeruli at the ultrastructural level.
Morphometrical investigation: The study included patients with g2 or g3 glomerulitis according to the Banff classification. Only biopsies with more than 10 glomerular profiles were included. If the patients had a previous biopsy with g0 score or a following biopsy with g0 these biopsies were included as well. On two sections from each patient, the mean glomerular volume was estimated by the method of point-sampled intercepts . The glomerular volume was estimated by the use of a computerized morphometrical system. The stereological method used is unbiased and the mean volume provided is the so-called volume-weighted mean glomerular volume . The total number of biopsies included were: 3 biopsies from patients with g0 prior to development of glomerulitis, 24 biopsies from patients with g2 or g3, and 14 biopsies from patients with g0 but who earlier had glomerulitis.
Electron microscopical investigation: Biopsies from 12 patients with g2 and g3 lesions had tissue were included for electron microscopy. Four biopsies without glomerulitis served as controls. A series of overlapping micrographs covering the whole glomerulus including Bowman's capsule was taken at an electron optical magnification of about 1000X. After enlargement these micrographs were glued together into a montage. Secondly, a series of 10 micrographs, systematically distributed over the glomerular cross section, was taken at 4000X. Each biopsy was qualitatively described with regard to prominent changes, which may include narrowing of capillary lumina, cell types and fibrin in the lumina, basement membrane deposits, mesangial interposition etc.
The ultrastructural investigation demonstrated that the biopsies with glomerulitis were more cellular than the controls. An important variation, however, existed within the glomerulitis group in that some glomeruli had their capillary lumens totally occluded by cells (Fig. 2), whereas others had open capillaries with scattered lymphocytes and macrophages. We also had the impression that the number of mesangial cells were increased. No immune deposits or humps could be found. Deposits of non-immune type were often seen and to an equal degree in the two groups. Especially deposits of type 3 (according to Bohman et al.) were frequently seen. A higher degree of foot process fusion was present in the ETG group, in some cases so pronounced that no splits were seen over a long distance (Fig. 2). Also the amount of smooth endoplasmatic reticulum in endothelial cells and less frequently in epithelial cells was increased (Fig. 2).
If glomerulitis has been present the graft survival is poor and transition to chronic graft glomerulopathy may occur [3, 10]. Spencer and Jepsen (1987) reported on 14 biopsies with early transplant glomerulitis collected from 1961 to 1980. The grafts with glomerulitis had poor 6 month graft survival: 14% against 42% in grafts without glomerulitis. In our investigation of 444 graft biopsies from the first 90 days post transplant, 13.5% of the biopsies (from 30 patients) had moderate or severe ETG (g2 or g3). Acute rejection and ETG were positively correlated but 40% of the biopsies with ETG had no acute rejection in the same biopsy. The graft survival in the patients with ETG was decreased but this seemed to be dependent on the simultaneous presence of acute rejection in the early post transplant course. Early transplant glomerulitis has been proposed to be a T cell-mediated allograft rejection where many of the intraglomerular T cells are CD8+11. The present ultrastructural study demonstrates that an immunological mechanism for glomerulitis must be different from immune complex deposition. As demonstrated the glomerular lesion is different from the classical endocapillary glomerulonephritis in the native kidneys. No immune deposits and no increase in the polymorph granulocytes were seen during ETG.
The glomerular volume is significantly increased during an episode of ETG and this is due to an increase in the number of cells and probably to a lesser degree due to the presence of swollen cells.
In conclusion, some of the changes are comparable to what is found in endocapillary glomerulonephritis in the native kidney but no immune deposits are found.