Penetration of intravesical doxorubicin in human bladders.

The bladder wall penetration kinetics of intravesical doxorubicin were examined in radical cystectomy patients, to provide insight into drug concentrations at target tumor sites. The dosing solution (40 mg/20 ml) was instilled just prior to the start of surgery and maintained for 60-115 min until just prior to bladder excision. The data showed considerable inter-patient variability in the peak plasma concentration (24-fold), urine concentration (7- fold), and tissue concentration (28-fold). The urine concentration at the time of tissue harvest was about 17% of the concentration in the dosing solution. This was due to the dilution by post-catheterization residual urine and urine produced during treatment. The doxorubicin concentration dropped by 32-fold across the urothelium, and declined semi-logarithmically with respect to depth in the capillary-perfused tissues beneath the urothelium with a 50% decrease over about 500 micromole. In three of six patients from whom tumor tissue was obtained, the doxorubicin concentration was higher than the adjacent non-tumor-bearing tissues of comparable tissue depth, whereas the reverse was seen in the remaining three tumors. The plasma concentrations were 0.02, 0.03, 0.05, 0.27, and 0.69% of the concentrations in the tumors, urothelium, lamina propria, superficial and deep muscle layers, respectively. These data indicate: (a) a considerable intra- and inter-patient variability in bladder tissue concentrations, in part due to the variability in the urine concentration; (b) the urothelium is an effective barrier to doxorubicin penetration; and (c) a targeting advantage of intravesical therapy for the treatment of superficial bladder cancer yielding superficial bladder tissue concentrations at least 2000-fold higher than in the systemic circulation. A comparison of the data of doxorubicin with our previously published data on mitomycin C shows similar bladder tissue pharmacokinetics for the two drugs, suggesting that there is no pharmacokinetic preference for either drug.