Clinico-pathological correlation of severe tubulointerstitial fibrosis in glomerular diseases.

ABSTRACT

BACKGROUND: Renal histopathology is the best method available to assess chronicity ofglomerular diseases. However renal biopsy is an invasive procedure and is available only in medical schools or tertiary-care hospitals in Thailand. Clinical predictors that discriminate the chronicity index of renal pathology may be valuable for the best timing of biopsy. The authors conducted this study to identify the clinical parameters of severe fibrosis in glomerular diseases. MATERIAL AND METHOD: The authors retrospective analyzed all consecutive patients with glomerular diseases who underwent ultrasound-guided renal biopsy in Siriraj Hospital between 2008 and 2010. The patients were statified according to degree of tubulointerstitial fibrosis (IF) into mild to moderate group (IF < 50%) and severe group (IF ≥ 50%). Data of clinical and radiological parameters which relate to severe fibrosis were obtained. Formula for prediction of advanced IF was also developed by backward stepwise logistic regression analysis. The authors also validated the model by application to the patients who underwent kidney biopsy in our center between 2011 and 2012. RESULTS: Of a total 682 patients, 169 patients (24.8%) were classified as a severe IF group. In the multivariate model, higher serum creatinine, lower mean length of both kidneys and systolic blood pressure (SBP) of more than 140 mmHg were significantly related to severe IF All three factors were incorporated into apredictive model e(x)/(1 +e(x)) while x = 1.3 + (0.6 x serum Cr in mg/dl)--(0.4 x mean length of both kidneys in cm)+(0.7 x 1 if SBP ≥ 140 mmHg or 0 if < 140 mmHg). The formula had AUROC of 0.82 and if calculated probability of fibrosis is higher than 0.37, it yields 90% specificity and 44% sensitivity for the prediction ofsevere fibrosis. When applied to 523 patients who underwent renal biopsy in 2011 and 2012, the sensitivity was 65.6% while specificity was 87.8%. CONCLUSION: High serum creatinine, presence of HT and decreased mean length of both kidneys are important clinical markers to predict renal fibrosis. The model constructed from these factors could be used in clinical practice for appropriate decision making.