Prednisolone resistance in childhood acute lymphoblastic leukemia: vitro-vivo correlations and cross-resistance to other drugs.

As an important determinant of response to chemotherapy, accurate measurement of cellular drug resistance may provide clinically relevant information. Our objectives in this study were to determine the relationship between in vitro resistance to prednisolone (PRD) measured with the colorimetric methyl-thiazol-tetrazolium (MTT) assay, and (1) short-term clinical response to systemic PRD monotherapy, (2) long-term clinical outcome after combination chemotherapy within all patients and within the subgroups of clinical good and poor responders to PRD, and (3) in vitro resistance to 12 other drugs in 166 children with newly diagnosed acute lymphoblastic leukemia (ALL). The 12 clinical poor PRD responders had ALL cells that were median 88-fold more in vitro resistant to PRD than 131 good responders (P = .013). Within all patients, increased in vitro resistance to PRD predicted a significantly worse long-term clinical outcome, at analyses with and without stratification for clinical PRD response, and at multivariate analysis (P </= .001). Within both the clinical good and poor responder subgroups, increased in vitro resistance to PRD was associated with a worse outcome, which was significant within the group of clinical good responders (P < .001). LC50 values, ie, lethal concentrations to 50% of ALL cells, for PRD and each other drug correlated significantly with those of all other 12 drugs, with an average correlation coefficient of 0.44 (standard deviation 0.05). The highest correlations were found between structurally related drugs. In conclusion, in vitro resistance to PRD was significantly related to the short-term and long-term clinical response to chemotherapy, the latter also within the subgroup of clinical good responders to PRD. There was a more general in vitro cross-resistance between anticancer drugs in childhood ALL, although drug-specific activities were recognized.