Background: Although several studies have dealt with clinicopathological correlations in diabetic glomerulosclerosis (DGS), few have investigated the prognostic value of the renal biopsy.
Methods: One hundred and thirty-five type2 diabetes mellitus (DM) patients with bioptically proven DGS and a mean follow-up of 56.1 months were subdivided in 5 groups according to the outcome: 1) living with preserved renal function; 2) living with renal failure, not requiring dialysis; 3) living in dialysis; 4) dead in dialysis; 5) dead with preserved renal function. Duration of DM, creatininemia and proteinuria values at the time of biopsy and a histologic scoring (from 0 to 3) of 10 morphologic features were considered for statistical analysis.
Results: Statistically significant differences were observed in the distribution of the above mentioned parameters in the 5 groups with the exception of the duration of DM. The prognosis of DGS is poor: 79 patients needed dialysis and 60 died. Univariate analysis demonstrated the prognostic value of creatininemia (>or= 1.5 mg/dL), proteinuria (>or= 3 g/d) and histologic score (>or= 10) in assessing the relative risk of progression to dialysis (OR= 9.75, 4.12 and 11 respectively). The prognostic value of the histologic score (<or= 1 vs >or= 2) was maintained when each morphologic parameter was separately evaluated (OR ranging from 2.8 to 8.5). Multivariate analysis was applied and only serum creatinine and histological score maintained their statistical significance (OR=4.45 and 2.46). The independence of these two variables means that the risk of progression to dialysis is multiplied in each single patient.
Conclusions: Renal biopsy adds reliable information to that supplied by clinical and laboratory findings in DGS and therefore its extensive adoption should be recommended. Thanks to the prognostic value of the single morphologic parameters, also small tissue samples can give reliable results.