RESUMEN
UNLABELLED: Study Type - Diagnostic (validating cohort). LEVEL OF EVIDENCE: 1b. What's known on the subject? and What does the study add? Microscopic haematuria (µH) is frequently detected in elderly adults. The American Urological Association recommends the follow-up of subjects with µH on bladder cancer. Whereas gross haematuria is considered an important sign of the presence of bladder cancer, the disease-predictive value of µH is less clear. No association of µH with the development of bladder tumours in a prospective screening cohort of chemical workers was observed. The positive predictive value of µH for bladder cancer was as low as 1.2%. Haematuria interfered with NMP22 but not with cytology and UroVysion(TM) test results. OBJECTIVE: ⢠To assess the positive predictive value (PPV) of microhaematuria (µH) and gross haematuria (GH) in bladder cancer screening and the influence of haematuria on tumour tests in a prospective study. PATIENTS AND METHODS: ⢠From September 2003 to January 2010, 1323 men took part in an annual voluntary bladder cancer screening programme for chemical workers with former exposure to aromatic amines. ⢠In 5315 urine samples haematuria was determined with a dipstick, followed by a microscopic blood cell count in the sediment. Haematuria was categorized into traces, µH and GH. ⢠Urinary leukocytes and other factors were investigated as potential predictors of haematuria using a generalized estimating equation model for repeated urinalysis. The risk of haematuria for positive tumour tests was analysed correspondingly. ⢠The bladder cancer risk was estimated for the highest degree of haematuria occurring during the study with Poisson regression. RESULTS: ⢠As of July 2010, 15 bladder tumours were detected in 14 participants. ⢠GH was found in four out of nine high-grade tumours and associated with a rate ratio of 3.82, 95% confidence interval (CI) 0.50-29.15 for the development of bladder lesions. ⢠The PPV of GH was 11.4%, but only 1.2% for µH. µH occurred in 18.8% of urine samples and was not associated with bladder cancer [rate ratio (RR) 0.72, 95% CI 0.11-4.78]. ⢠Abundant urinary leukocytes were associated with µH [odds ratio (OR) 8.34, 95% CI 2.26-30.69] and even stronger with GH (OR 22.25, 95% CI 6.42-77.06). ⢠Haematuria and leukocytes influenced NMP22 positivity (µH: OR 1.63, 95% CI 1.06-2.51, abundant leukocytes: OR 8.90, 95% CI 1.58-50.16), but not test results for urine cytology and UroVysion(TM) . CONCLUSION: ⢠While the PPV of µH for bladder cancer was low, there was a strong influence of haematuria and leukocytes on the protein-based tumour test NMP22®. ⢠Erythrocytes and leukocytes should be determined at least semi-quantitatively for the interpretation of positive NMP22 test results. ⢠In addition, a panel of tumour tests that includes methods not affected by the presence of erythrocytes or leukocytes such as cytology and UroVysion(TM) would improve bladder cancer screening.