Several conclusions can be drawn from a review of HMM preclinical and clinical pharmacology data. The drug is extensively metabolized by animals and by man. The drug is well absorbed following oral administration to animals, but oral bioavailability is low due to first pass metabolism. Based on limited human data and more complete animal data, absorption of HMM following oral administration may be quite high in man. We do not yet know the oral bioavailability of HMM in patients, but again based primarily on animal studies, oral bioavailability is most likely low and variable due to extensive first pass metabolism. Systemic exposure to HMM and demethylated metabolites following oral administration varies greatly from patient to patient and is sometimes quite low. Most patients are, however, exposed to a substantial fraction of the administered dose when determined by urinary recovery of the total dose (based on parent drug and metabolites or total radioactivity) or by the total plasma AUC of parent drug and all metabolites. Systemic exposure to HMM following intravenous administration is clearly greater and less variable than following oral administration. An unresolved question is whether the highly variable and often low systemic exposure after oral administration compromise antitumor activity when compared to intravenous administration. A key issue is whether or not one accepts the hypothesis that metabolism is a prerequisite for antitumor activity. The metabolic activation studies do not rule out other mechanisms of HMM antitumor activity. Modest activity of HMM was observed after prolonged exposure to cells which did not metabolize the drug (38, 48). However, most of the accumulated data are consistent with the metabolic activation hypothesis. Certainly HMM has clinical activity when administered by mouth. If metabolism is required, then exposure to the total dose (parent drug and metabolites) could be of significance even when exposure to HMM is low, since every demethylated metabolite must have come ultimately from the initial HMM demethylation. We do not know whether the initial metabolic reaction (occurring in the liver rather than in the tumor) provides sufficient exposure of tumor to reactive species. Specifically, does the variable HMM plasma AUC seen after oral administration lead to variable delivery of potentially reactive species to tumor (by rapid breakdown and/or further metabolism of MPMM before it leaves the gut and/or liver) or are quantities of MPMM delivered to tumor comparable to those delivered following intravenous administration. The issue of rate of MPMM formation compared to rate of breakdown and ultimate delivery to tumor has been noted by Judson and Rutty (31). As they point out, pharmacokinetic studies characterizing HMM and MPMM plasma elimination following oral and intravenous administration would provide important data in this regard. Pharmacology studies with HMM will continue to provide data important in better understanding the mechanism of action of this unique agent and to assist in its better use in the treatment of malignancies. © 1991.