Recurrent SA7HD leukemia cells have also "lost" the response to stimulation by L-929-conditioned medium (L-929-CM) and indeed were significantly insensitive to proliferative stimulation by combinations of WEHI-conditioned medium (WEHI-CM) and L-929-CM (P = 0.04) in contrast to untreated SA7HD leukemia, which displayed dose-dependent stimulation by L-929-CM and exhibited synergistic proliferative response to combinations of the two growth factors. This altered growth-factor sensitivity, which was a function of both the route of leukemia inoculation and drug administration, was detected by both the colony and the tritiated thymidine (3HTdR) incorporation assays. In contrast to transplant in normal mice, recurrent leukemia transplanted in mitoxantrone pretreated mice remained growth-factor insensitive. In addition, no growth delay of the recurrent leukemia occurred during this passage in pretreated mice. However, recovery of growth-factor sensitivity still occurred after a single subsequent transplant in normal mice. The magnitude of resistance developed by the recurrent leukemia was not increased by passage in mitoxantrone pretreated mice. Rather, the extent of resistance was a function of the dose of mitoxantrone employed in the treatment of the leukemia before recurrence. These data interpreted along with previously published observations suggest that the development of altered growth-factor sensitivity by SA7HD recurrent leukemia was directly linked to the presence of mitoxantrone. In addition, this phenomenon possibly confers survival advantage to the leukemia cells and this probably accounts for the observation that most of the drug-treated mice ultimately developed recurrent disease despite adequate treatment.