Clinical Evaluation of the miR Sentinel BCa™ Test to Diagnose Bladder Cancer in Hematuria Patients

Primary Objective(s):

Establish the performance characteristics of the miR Sentinel™ BCa Test as an aid in diagnosing bladder cancer in men and women presenting with hematuria.

Secondary Objective(s) Establish the performance characteristics of the miR Sentinel™ BCR Test in identifying bladder cancer patients with recurrent disease.

Study Design:

Participants who present with evidence of hematuria (micro- or macro-hematuria) and are scheduled to have cystoscopy will be enrolled. All participants will provide a urine sample for the miR Sentinel™ BCa Test prior to cystoscopy for bladder cancer. The presence or absence of cancer will be determined by cystoscopy. Diagnosis of bladder cancer will be confirmed by histopathology. The performance of the miR Sentinel™ BCa Test for classifying future patients as either bladder cancer or no bladder cancer will be established. The sensitivity, specificity, positive and negative predictive value determined for the miR Sentinel™ BCa Test and will be compared to cystoscopy.

Statistical Analysis:

Definitions. The general notation P(A | B) means the probability (prevalence) of the event A among the subgroup of patients in a population all of whom have the event B.

1. Sensitivity = P(screening test + | disease).
2. Specificity = P(screening test - | no disease).
3. PPV = P(disease | screening test + ).
4. NPV = P(no disease | screening test -).

Estimate of Sample Size. This study is targeted to enroll 3000 participants, 1500 males and 1500 females, who present with hematuria. Approximately 15% of participants with hematuria have bladder cancer; men have a higher incidence of bladder cancer than women. Therefore, the investigators anticipate that 225 males (15% of all males enrolled) and 175 females (approximately 12% of all women enrolled) will have bladder cancer confirmed by histopathology.

Overview. As outlined below, this clinical study will enroll male and female participants that have been identified as having (micro- or macro-) hematuria and are therefore at risk for bladder cancer and meet clear eligibility criteria. The study will evaluate the miR Sentinel™ BCa Test for classification of participants as cancer or no cancer, to be used as an aid in diagnosing bladder cancer. For participants diagnosed with bladder cancer who undergo TURBT will be eligible to continue in the study of miR Sentinel™ BCR Test to monitor for recurrent disease in conjunction with the cystoscopy associated with the standard of care recommended by the AUA.

The miR Sentinel™ BCa Test functions by controlling sensitivity at or above a pre-specified level, denoted 1-α; where α is the false-negative rate of the test (a patient who truly has bladder cancer has a negative test result), For example, the value that has been assumed in this design is α≤0.05 so that sensitivity is at least 95%. To describe how the cutoff for the miR Sentinel™ BCa Test is calculated to control sensitivity, for each participant determined to have bladder cancer by histopathology, the miR Sentinel™ BCa Score will be calculated using an approach that blinds each participant's true cancer status, mimicking the setting for classification of a future patient. The cutoff value for the Sentinel™ Score is then chosen so that the empirical sensitivity, calculated over all patients in the training dataset, yields a 95% upper confidence interval with lower limit 1-α.

To provide some detail on the calculation of the error rates of the test, the method to be used for estimation is cross-validation. Cross-validation produces unbiased estimators of error rates for classification of a future patient, with unknown true disease status, as the number of patients in the training dataset grows large. If there are 3000 participants with hematuria divided equally from male and female population in the study and 225 males and 175 females with bladder cancer, then, for males, the upper 95% confidence interval based on an observed empirical sensitivity of 97% (218/225) will have a lower bound of 0.95, and for females, the upper 95% confidence interval based on an observed empirical sensitivity of 98% (171/175) will have a lower bound of 0.95.

Validation of the miR Sentinel™ BCR for identifying patients with recurrent disease uses essentially the same statistical methodology. However, the classification is based on the comparison of the miR Sentinel™ BCR signature for patients with stable (non-recurrent disease) and patients with evidence of progressive (recurrent) disease. The longitudinal monitoring of participants diagnosed with bladder cancer by cystoscopy may provide additional information related to anticipatory changes in the Sentinel™ signature indicative of recurrence.