Background: The objective of this study was to test the hypothesis that elevated urinary levels of soluble Fas (sFas) would aid in the surveillance of patients with a past history of nonmuscle-invasive transitional cell carcinoma (TCC) of the urinary bladder.
Methods: sFas levels were determined in cell lysates and supernatants from 2 human bladder cancer cell lines (T24 and TCC-SUP) and in voided urine from 188 consecutive patients who were at risk for TCC recurrence, 31 patients who had noncancerous urologic conditions, and 10 healthy individuals. The authors also obtained barbotage cytology and voided nuclear matrix protein 22 (NMP22) levels. sFas was analyzed continuously and categorically on the basis of its quintile distribution.
Results: sFas was present in cell lysates and conditioned media from both cell lines. sFas levels were found to be higher in the TCC group (n = 122 patients) compared with the control group (P < .001). Higher levels of sFas were associated with positive cytology assay results (P < .001), higher NMP22 levels (P < .001), NMP22 levels > 10 U/mL (P < .001), and tumor stage > or = T1 (P < .001). The areas under the receiver operating characteristics (ROC) curves of sFas and NMP22 for bladder cancer detection were 0.757 (95% confidence interval, 0.694-0.819) and 0.704 (95% confidence interval, 0.637-0.772), respectively. In the > 75% sensitivity region of the ROC curves, sFas was consistently more specific than NMP22. In multivariate analyses, sFas, NMP22, and cytology all were found to be associated with the presence of bladder cancer (P values < or = .009), but only sFas and cytology were associated with tumor stage > or = T1 (P values < or = .026).
Conclusions: sFas was produced and released by bladder TCC cells. Urine sFas was an independent predictor of bladder cancer recurrence and invasiveness in patients who had a past history of nonmuscle invasive bladder TCC, and it outperformed NMP22.
2006 American Cancer Society