In idiopathic membranous nephropathy (MN), the main predictors for progression to chronic renal failure (CRF) are the amount of proteinuria and extent of tubulointerstitial damage. The aim of this study is to evaluate whether urinary excretion of proteins reflecting the alteration of permselectivity in the glomerular capillary wall, such as immunoglobulin G (IgG), and the reabsorption impairment of low-molecular-weight proteins, such as alpha(1)-microglobulin (alpha(1)m), correlates with the extent of tubulointerstitial damage and have a predictive value for functional outcome and response to therapy better than 24-hour proteinuria. In 78 patients with MN, urinary excretion of albumin, transferrin, IgG, and alpha(1)m was measured by immunonephelometry in second-morning urine samples and expressed in milligrams per gram of urinary creatinine (uCr). In 48 patients with characterization of proteinuria and renal biopsy performed at the same time, excretion of IgG (P = 0.0087) and alpha(1)m (P = 0.0024), but not albumin (P = 0.37), transferrin (P = 0.38), or 24-hour proteinuria (P = 0.32), was associated significantly with the extent of tubulointerstitial damage (score, 0 to 1 versus >/=2). Only alpha(1)m excretion was associated significantly with global glomerular sclerosis (P = 0.0032) and arteriolar hyalinosis (P = 0.0004). Moreover, urinary excretion of alpha(1)m was significantly dependent on IgG excretion (r = 0.67; P = 0.0001), but not on albumin (P = 0.66) or 24-hour proteinuria (P = 0.07). Functional outcome could be evaluated in 38 patients with nephrotic syndrome and baseline normal renal function (serum creatinine, 0.99 +/- 0.20 mg/dL; follow-up, 44 +/- 22 months). Remission was 100% versus 20% in patients with IgG excretion less than 110 mg/g uCr versus 110 mg/g uCr or greater (P = 0.0001) and 77% versus 17% in patients with alpha(1)m excretion less than 33.5 mg/g uCr versus 33.5 mg/g uCr or greater (P = 0.0009), respectively. In patients with IgG and alpha(1)m excretion less than or greater than the cutoff value, progression to CRF was 0% versus 35% (P = 0.0026) and 0% versus 58% (P = 0.0001), respectively. Nineteen patients treated with immunosuppressive therapy were compared with 19 untreated patients. There was no difference in remission or progression between treated and untreated patients when IgG and alpha(1)m excretion were less than the cutoff value. There was a significant difference for progression to CRF between treated and untreated patients when alpha(1)m excretion was greater than the cutoff value (17% versus 100%; P = 0.0076). In conclusion, IgG excretion is associated significantly with the extent of tubulointerstitial damage and alpha(1)m excretion. This observation supports the hypothesis that IgG may be the toxic moiety of proteinuria. Excretion of IgG and alpha(1)m has a significant predictive value for both remission and progression and is useful to identify patients who are at risk for progression and for whom treatment with immunosuppressive therapy is indicated soon after diagnosis.