Predictors of response of acute nonlymphocytic leukemia

Given the time and effort expended by investigators and given the large numbers of patients studied, there are disappointingly few "predictors" of response. From the clinical perspective, aside from performance status, age, prior history of toxic exposure, and grossly abnormal organ function, there are essentially no reliable indicators of the likelihood of a patient surviving remission induction therapy. The absence of such indicators might reflect the fact that without grossly abnormal organ function, all patients generally begin with an equal possibility of survival and that events which occur during therapy determine survival. One thing is certain, death directly attributable to leukemic cell overgrowth despite chemotherapy is an extremely rare event. Hence, therapeutic inadequacy in this sense, at the time of initial diagnosis, is not a common cause of treatment failure. Studies aimed at the prediction of leukemic cell responsiveness to therapy have been plagued by two general problems. The first is that the drug sensitivity assays have been quite primitive. For example, the ability of a cell to take up a drug is not synonymous with sensitivity to that drug. Additionally, tests which are dependent upon assays capable of making measurements in only a small subpopulation of cells, such as in vitro clonogenic assays, are likely to have only limited applicability. On the other hand, assays which measure the properties of the leukemic cell population as a whole are incapable of recognizing arabinoside therapy have provided data which suggest that in addition to patient survival three conditions must be satisfied if a complete remission is to occur: the pretherapy leukemic cell mass must be moderate or low, an adequate number of cells must be synthesizing DNA, and cytosine arabinoside must produce significant inhibition of DNA synthesis in vitro. Each factor is consistent with what is known about cytosine arabinoside: it is an S-phase-specific agent which must be incorporated into DNA in order to kill leukemic cells. When the relationship between these same factors and response to combination chemotherapy were studied, not unexpectedly, no relationship was discerned. Finally, in this setting pharmacokinetic studies have demonstrated that the amount of araCTP formed in leukemic cells in vivo when doses of 2 g and 3 g/m2 are administered are indistinguishable, thereby explaining the clinical equivalence of these two dosage levels.(ABSTRACT TRUNCATED AT 400 WORDS)