Identification of non-diabetic glomerular disease in renal biopsies from diabetics—a dilemma

diabetic renal disease has augmented tremendously during the last half century [1,2] and during the last abnormalities. decade we have been able to understand at least some elements of its pathogenesis. Nephropathy in a patient Glomerular lesions of diabetic nephropathy suVering from DM may, however, not be related directly to the diabetic disease but may be a complicat- Following the first seminal report by Kimmelstiel and ing renal disease, not due to the diabetic metabolic Wilson [8] and later on worked out in great detail by abnormality. The situation is further complicated since others, we have now a clear idea of the histopathology logically both lesions may be present in the same of diabetic renal disease and its progress with time. patient. While it was formerly held that non-diabetic We know now, that at the onset of insulin dependent renal disease was of minor importance in DM, this DM (IDDM ) (in this type the zero time is well defined) seems not so obvious to-day, and opinions of investig- there is enlargement of the glomeruli indicating ators about prevalence of non-diabetic renal disease in increased glomerular function, but no other structural DM, as revealed by renal biopsies, have been con- lesions by light microscopy (LM ) or electron microflicting. The correct diagnosis is important for the scopy (EM ). The first ultrastructural lesions (widening patient since prognosis and treatment may vary accord- of the mesangial matrix and thickening of the periing to the underlying cause. Furthermore: if a substan- pheral glomerular basement membrane) can be mortial part of renal disease in diabetics had other causes phometrically identified 2‐3 years after onset [9]. than the diabetic metabolic derangement, this must be Changes can usually not be seen by LM until several a confounding factor in the interpretation of interven- years of diabetic disease. They are described in two tion trials. forms. In the diVuse form there is a widening of all The classical diabetic nephropathy is morpholo- mesangial regions as well as thickening of the perigically primarily a glomerulopathy, characterized clin- pheral capillary basement membranes, corresponding ically by proteinuria and progressive decrease of to what is already apparent at an earlier time from glomerular filtration. Tubular atrophy and interstitial morphometry on EM pictures. The diVuse lesion is fibrosis is a common final pathway for many renal most conspicuous in PAS-stained sections. In the nodudiseases including diabetic nephropathy. These lesions lar type some of the mesangial regions take on the have formerly been interpreted as ischaemic sequels to appearance of round nodules which are acellular in vascular and glomerular lesions but they may also their centres but surrounded by a rim of mesangial occur in diabetics with only mild vascular and glomer- nuclei. While these types are sometimes described as ular lesions and at an early stage of diabetic renal separate histologic entities, there are good reasons to disease. In some reports this was even rather frequent regard the nodular type as a further development of [3,4] and it is today not at all obvious that tubulointer- the diVuse type. Missing a formal proof of this (e.g. stitial lesions are of ischaemic origin. A competing by the study of sequential biopsies from the same hypothesis relates them to the diabetic state, just as patients), it is clear that nodules are formed from mesangial matrix, that they occur later in the disease