The correlation of maximal drug dose, tumor recruitment, and sequence timing with therapeutic advantage: Schedule‐dependent toxicity of cytosine arabinoside

Studies compared the induction of maximal tumor regrowth and thereby sensitivity to a cell cycle‐specific agent, and the relative cytotoxic effects on stimulated normal host tissue growth in the LBN rat bearing AML. The maximal tolerated dose (MTD) of Ara‐C (1800 mg/m² total dose) given in a timed sequence produced significant antileukemic effects, but provided a narrow therapeutic index. The initial dose of Ara‐C, 100 mg/m² q 8 hr×6 sub q followed by a second dose, 50 mg/m² q 8 hr×6, given at varied intervals, produced the greatest antitumor effect when given on days 0, 1‐6, 7, while the continued sequence (0‐1,2‐3) was toxic, with the MTD far less (600 mg/m²). When the discontinuous sequence was given at the time of GI mucosa recovery (0,1‐4,5) less than 1/6 MTD could be administered without death. In contrast, when the 2nd dose was given later (8‐9,10‐11, 12‐13) shorter remission durations were achieved even at > MTD of Ara‐C than the 0‐1, 6‐7 schedule. The discontinuous optimally timed schedule allowed for regrowth and then maturation of normal host tissues (GI) out of synchrony with the continued proliferation of residual tumor. A schema for prevention of Ara‐C toxicity in man based on induced changes in normal and leukemic cell growth, which combines timing of drug with gut rest, hyperalimentation, and diuresis, has been clinically effective.