No common final pathogenetic pathway in haemolytic uraemic syndromes.

There is a long standing and widely held view that the many sub-forms of the haemolytic uraemic syndrome (HUS) and thrombotic thrombocytopenic purpura (TTP) share a common final pathogenetic pathway. This was expressed recently in a review article in this journal [1]. This is an assumption which deserves to be challenged. It appears to be based on the concept that the histological hallmark of these syndromes, thrombotic microangiopathy (TMA), is a single entity. There is growing evidence that it is not. Symmers originally used the term TMA to describe thrombi occurring in an arteriolar distribution, particularly at the junction between arteriole and capillary, in the absence of vasculitis [2]. Later, Habib found such lesions in infants with HUS and subsequently described different patterns of TMA, expanding the term to cover all of them. One form, remaining close to Symmers’ description, consisted of lesions at the arteriole with or without extension into the glomerular capillaries. In a second type, thrombosis was confined to the glomerulus, with endothelial swelling and amorphous material in the subendothelial space. The third form affected larger arteries as well as arterioles, and glomeruli appeared ischaemic. Various correlations were made between the histological sub-types and the clinical features of patients, and an extensive and valuable summary of this work appeared in 1992 [3]. In brief, arterial lesions were more often associated with severe hypertension and adverse outcome, while glomerular thrombosis alone was more favourable, as long as there was no co-existing extensive renal cortical Fig. 1. Necrotic kidney from a child with VTEC-induced HUS. A glomerulus is distended by thromboses, with a thrombus in the necrosis [4]. Others have reinforced this view. afferent arteriole. Periodic acid–methenamine silver (PAAg). While the histological sub-type of TMA could be related to prognosis, early studies were not in a position the numerically most important group. Furthermore, to match this to the clinical groups discernable today. it was only at the beginning of that decade that a For example, much of this work was performed before prodromal illness of diarrhoea was shown to have the mid-1980s, the time when the commonest single prognostic significance, allowing a broad interim classicause of HUS became recognized, i.e. infection by fication into diarrhoea-associated (D+) and nonverocytotoxin (shiga-like toxin)-producing Escherichia diarrhoeal forms [5]. In Europe and the Americas, coli (VTEC). This precluded histological definition in there is a close association between D+HUS and VTEC infection in children. On the other hand, nonCorrespondence and offprint requests to: Dr C. M. Taylor, diarrhoeal HUS in adults or children is a heterogeneous Department of Nephrology, The Birmingham Children’s Hospital, Steelhouse Lane, Birmingham B4 6NH, UK. collection of rare diseases. Only a minority of patients