Value of multiparametric magnetic resonance imaging for local staging of invasive urinary bladder tumours.

Background: Initial tumour staging in bladder cancer mainly relies on the histo-pathological outcome of the transurethral bladder tumour resection (TURBT) and imaging by means of a CT-scan (CT-intravenous urography; CT-IVU). The reported risk of understaging varies from 24-50%. To further improve the the evaluation of depth of invasion of the bladder tumour the application of magnetic resonance imaging (MRI) may be useful. To substantiate the additional value of this imaging modality the present observational study was designed. Study design: This is a prospective observational study to analyse bladder tumour staging with multiparametric magnetic resonance imaging (mpMRI) in patients with a known bladder tumour, who are planned for radical cystectomy. Study population: Patients with an invasive bladder cancer who are planned for radical cystectomy. Intervention: Patients were accrued during their visit to the outpatient department of urology. They underwent routine cystoscopy, laboratory tests (including serum Creatinin) and CT-IVU investigations and subsequently a mpMRI. Main study parameters/endpoints: To demonstrate the value of mpMRI in the initial staging of bladder tumours using radiological bladder tumour stage (T-stage) based on mpMRI and pathological bladder tumour stage based on ‘whole-mount’ histo-pathology after radical cystectomy. Results: Thirty-seven participants with known bladder tumours underwent mpMRI and subsequent cystectomy. After mpMRI 10 participants were diagnosed with non-muscle-invasive bladder cancer (NMIBC) and 27 participants with muscle-invasive bladder cancer (MIBC). In the ‘whole-mount’ pathology results 12 participants had NMIBC and 25 participants had MIBC. We found a sensitivity and specificity of 0.88 en 0.58 respectively, for the evaluation of MIBC. The positive and negative predictive value were 81% and 70% respectively. The diagnostic accuracy of mpMRI to differentiate between NMIBC and MIBC was 78%. Conclusions: We found a sensitivity of 88% and a specificity of 58% for mpMRI to discriminate NMIBC from MIBC.