This study was performed to evaluate the clinical usefulness of nuclear matrix protein 22 (NMP22) as urinary marker for bladder cancer and to define its role in comparison with urinary cytology in diagnostic management of patients with hematuria.
Materials and methods: NMP22 values in voided urines were determined on 90 patients (81 males, 14 females) with macro or microscopical hematuria, using the NMP22 test-kit (Matritech) based on an enzyme-linked immuno-sorbant assay. The cut-off value for positive samples was 10U/ml. In all cases urinary cytology was performed on the same sample. Patients suspected for the presence of a transitional cell carcinoma (TCC) underwent cytoscopic control. Statistical signification of the medians difference was analyzed using both medians variance analysis and Student's test. The increasing in diagnostic predictivity was analyzed elaborating contingency tables and performing consequently chi 2 test.
Results: 32 cases of TCC were endoscopically detected: sensitivity of cytopathology was 75.8%, specificity was 62.5%. The positive predictive value of the cytology was 22% and negative predictive value was 49%. The results concerning NMP22 dosage are: sensitivity 84%, specificity 62%. Positive predictive value of NMP22 test was 30% and predictive negative value was 40%. No significant differences between cytopathology and NMP22 dosage were founded (p > 0.995). Considering both cytopathology and NMP22 dosage, sensitivity was 82.7% and specificity raises up to 90.6% with statistical signification (P < 0.001). The median NMP22 value in patients affected by TCC endoscopically confirmed (group A) was 55.2 U/ml, in subjects with no evidence of malignancy (group B) it was 19.1 U/ml. The difference shows statistical signification (p < 0.001). In 20 cases with TCC, hystological grading were available and were investigated in relationship with NMP22 title in U/ml, median NMP22 value for each grade was: CIS = 102 U/ml, G0 = 35 U/ml, G1 = 30 U/ml, G2 = 66 U/ml, G3 = 54 U/ml. It can be assessed that NMP22 test is useful in differentiating subjects affected by TCC from subjects with no evidence of malignancy and may help in early diagnosis of TCC and in predicting recurrent even if low grade tumors especially if used in association with cytology and endoscopy.