Pharmacodynamics of doxorubicin in human bladder tumors

Intravesical doxorubicin treatment delivers high drug concentrations to the bladder wall, yet the treatment produces only a variable and incomplete response in superficial bladder cancer and insignificant activity in muscle-invading disease. This study evaluated the pharmacological basis for the clinical observations and potential prognostic indicators of tumor sensitivity to doxorubicin. The pharmacodynamics of doxorubicin were studied using histocultures of surgical specimens of seven superficial (Ta and T1) and nine invasive (T2-T4) tumors. After a 2-h exposure, drug treatment caused a concentration-dependent inhibition of proliferation, as measured by bromodeoxyuridine incorporation into DNA. Extensive cell death was observed at high concentrations (>10 micrometer) in sensitive tumors. The IC50 values ranged from 0.14 to 5.2 micrometer for superficial tumors and from 1.4 to >100 micrometer for invasive tumors. A comparison of the IC50 and IC90 values with the drug concentrations previously determined in human bladder walls showed that IC50 was achieved in all Ta tumors, 67-100% of T1 tumors, and one of the nine invasive tumors, whereas IC90 was achieved in all Ta tumors but none of the other tumors. To determine the biological basis of variable doxorubicin sensitivity among different tumors, we examined the relationship of chemosensitivity with tumor pathology and with expression of multidrug resistance p-glycoprotein (Pgp) and p53 protein. The invasive, high-grade, highly proliferative, p53- and Pgp-positive tumors were more resistant than the superficial, lower-grade, p53- and Pgp-negative tumors. All tumors were negative for bcl-2. The rank order of these factors was p53 expression > tumor stage > grade > bromodeoxyuridine labeling index > Pgp. Statistical analysis using the Akaike Information Criterion indicates that the two-parameter combination of p53 expression with stage further improved the predictive value. The present study shows that: (a) there was a >700-fold difference in doxorubicin sensitivity among superficial and invasive tumors; (b) the variable and incomplete response of superficial bladder cancer to intravesical doxorubicin therapy is likely due to the 35-fold variability in tumor chemosensitivity and, to a lesser degree, the 4-fold variability in tissue pharmacokinetics; (c) the lack of response of invasive cancer to intravesical doxorubicin therapy is likely the result of the inadequate drug concentrations presented to the deep muscle layers and the low chemosensitivity of the more aggressive tumors; and (d) the combination of p53 expression and high stage was the most significant predictor of doxorubicin sensitivity, whereas Pgp expression was the least important.