Dr. Gossmann suggests that our finding of an association between nonglomerular hematuria and bladder cancer in patients receiving cyclophosphamide may be biased because we did cystoscopy only in patients with nonglomerular hematuria. Dr. Gossman does not accurately quote the literature when stating that 25% of cases of bladder cancer are not associated with hematuria. The reference actually states that as many as 25% of patients with bladder cancer may not have hematuria as their initial clinical manifestation [1]. Such patients have other signs and symptoms of tumor invasion that lead to diagnostic cystoscopy, and nearly all patients with bladder cancer develop hematuria or abnormal urine cytologic findings during their illness.

The patients in our study were seen every 6 to 12 months after cyclophosphamide therapy was discontinued so that we could monitor for relapse of disease and for long-term complications of cyclophosphamide therapy. This monitoring included a thorough history and physical examination, routine urinalysis, and routine examination of urine cytologic findings. The median duration of follow-up for our patients was 8.5 years, and the investigators involved in the study were aware of the potential urotoxicity of cyclophosphamide. Any patient who developed bladder cancer and did not initially manifest hematuria would show other signs and symptoms of tumor invasion, eventually develop hematuria, or develop positive urine cytologic findings. Any of these manifestations of bladder cancer would have been detected during our long-term follow-up evaluations and would have led to a cystoscopic examination. This point is shown by two of our patients who did not have hematuria when their cancer was discovered. These patients had cystoscopy for evaluation of abnormal urine cytologic findings that had been identified during routine follow-up evaluations. Their episodes of nonglomerular hematuria occurred 4 and 5 years before the diagnosis of bladder cancer. Thus, it is unlikely that we missed many cases of bladder cancer in patients without nonglomerular hematuria. Our study documents that the occurrence of nonglomerular hematuria identifies a subgroup of cyclophosphamide recipients who are at high risk for the subsequent development of bladder cancer.