Combination chemotherapy: arriving at optimal treatment levels by incorporating side effect constraints.

This paper represents a method for determining optimal dose levels in a drug combination by analyzing the dose-response surfaces estimated from experimental data. In addition to using the presence or absence of a favorable response (in this instance, survival for at least twice the median untreated lifespan of B6D2F1 female mice bearing advanced sc implanted L1210 leukemia), the presence or absence of associated side effects is also employed (in this instance, Day-12 weight loss). The technique is potentially superior to existing methods of responding to the occurrence of treatment side effects due to its ability to take into account interactions between drugs. Since this method offers a way to determine optimal dose by including nonlethal toxicity information; it should find application in clinical situations.