AIM The usefulness of renal pathologic diagnosis in type II DM (diabetes mellitus) remains debate.

METHODS We grouped the pathologic diagnoses as pure DN (diabetic nephropathy), NDRD (non-diabetic renal disease), and NDRD mixed with DN (Mixed). We classified pure DN as the criteria suggested by Tervaert. We compared the accuracy of clinical parameters to predict DN and usefulness of pathology to predict renal prognosis.

RESULTS Among 126 enrolled patients, there were 50 pure DN, 65 NDRN, and 11 Mixed. The sensitivity and specificity for predicting DN with the presence of retinopathy were 77.8-73.6% and, with a cut-off value of 7.5 years of diabetic duration, the sensitivity and specificity were 64.5-67.2%. ESRD (end stage renal disease) occurred in 44.0% of DN, 18.2% of Mixed, and 12.3% of NDRD (p<0.001). Among pure DN, Class IV showed the lowest estimated glomerular filtration rate (eGFR). We estimated the 5-year renal survival rate as 100.0% in Classes I and IIa, 75.0% in Class IIb, 66.7% in Class III, and 38.1% in Class IV (p=0.002).

CONCLUSIONS Nephropathy of type II DM was diverse and could not be completely predicted by clinical parameters. The renal pathologic diagnosis was a good predictor for renal prognosis in type II DM.