Microsoft Word - NEF081BF

G. Rostoker, Department of Nephrology, INSERM U 139 and Fond d’Etudes du Corps médical H. Mondor, F-94010 Créteil (France) Sir, A possible relationship of primary IgA nephropathy (IgA GN) with wheat gliadin and latent celiac disease (CD) has been suggested by the presence of high levels of circulating IgA antigliadin antibodies [5], the decrease of IgA-containing circulating immune complexes after gluten-free diet [3] and the few cases of IgA GN associated with patent CD and/or dermatitis herpetiformis (DH) [4]. Antireticulin and IgA class antiendomysium antibodies (IgA EmA) are specific immunological markers for CD and DH associated with CD [2,6]. The IgA EmA are found only in gluten-sensitive enteropathy and their formation precedes the development of the pathological gut changes in CD and DH [1]; these autoantibodies become undetec-table in CD when the flattening of the gut mucosa disappears after gluten-free diet [1]. Therefore, we looked for antireticulin and IgA EmA in the sera of IgA GN patients with high levels of circulating IgA antigliadin antibodies. We analyzed the sera of 19 untreated patients with IgA GN having permanent proteinuria, according to the usual his-tological and immunofluorescence criteria [4], ( > 1 < 3 g/24 h; age: mean 41 years, range 24–58). All the patients exhibited high levels of circulating IgA antigliadin antibodies (ELISA technique DO > 0.65; mean DO: 1.11, range 0.66– 2.6 [5]); they were free of abdominal pain, diarrhea and malabsorption symptoms. All the patients were on a normal diet. The antireticulin and antiendomysium antibodies were studied with indirect immunofluorescence technique with sera tested at the dilutions of one fifth and one tenth; for antireticulin antibodies the substrates were unfixed frozen sections of rat liver, kidney and gastric mucosa [6] and for antiendomysium antibodies the substrates were frozen sections of monkey esophagus (Zeus-Scientific Inc.) [2]. The conjugates were fluores-cein-conjugated antibodies to human immunoglobulins (IgG, IgA, IgM: Diagnostic Pasteur, dilution 1/100) and to human α-chain (Dako, dilution 1/20). No patients exhibited either the specific Rl pattern for antireticulin antibodies or IgA antiendomysium. Therefore, the lack of antireticulin and IgA EmA specific to glutensensitive enteropathy in 19 patients with IgA GN and IgA antigliadin antibodies would tend to