Patients with disseminated Ewing's family of tumors (ESFT) often experience drug-resistant relapse. We hypothesize that targeting minimal residual disease with the cytotoxic retinoid N-(4-hydroxyphenyl) retinamide (4-HPR; fenretinide) may decrease relapse. We determined the following: (a) 4-HPR cytotoxicity against 12 ESFT cell lines in vitro; (b) whether 4-HPR increased ceramide species (saturated and desaturated ceramides); (c) whether physiological hypoxia (2% O(2)) affected cytotoxicity, mitochondrial membrane potential (DeltaPsi(m)) change, or ceramide species or reactive oxygen species levels; (d) whether cytotoxicity was enhanced by l-threo-dihydrosphingosine (safingol); (e) whether physiological hypoxia increased acid ceramidase (AC) expression; and (f) the effect of the AC inhibitor N-oleoyl-ethanolamine (NOE) on cytotoxicity and ceramide species. Ceramide species were quantified by thin-layer chromatography and scintillography. Cytotoxicity was measured by a fluorescence-based assay using digital imaging microscopy (DIMSCAN). Gene expression profiling was performed by oligonucleotide array analysis. We observed, in 12 cell lines tested in normoxia (20% O(2)), that the mean 4-HPR LC(99) (the drug concentration lethal to 99% of cells) = 6.1 +/- 5.4 microm (range, 1.7-21.8 microm); safingol (1-3 microm) synergistically increased 4-HPR cytotoxicity and reduced the mean 4-HPR LC(99) to 3.2 +/- 1.7 microm (range, 2.0-8.0 microm; combination index < 1). 4-HPR increased ceramide species in the three cell lines tested (up to 9-fold; P < 0.05). Hypoxia (2% O(2)) reduced ceramide species increase, DeltaPsi(m) loss, reactive oxygen species increase (P < 0.05), and 4-HPR cytotoxicity (P = 0.05; 4-HPR LC(99), 19.7 +/- 23.9 microm; range, 2.3-91.4). However, hypoxia affected 4-HPR + safingol cytotoxicity to a lesser extent (P = 0.04; 4-HPR LC(99), 4.9 +/- 2.3 microm; range, 2.0-8.2). Hypoxia increased AC RNA expression; the AC inhibitor NOE enhanced 4-HPR-induced ceramide species increase and cytotoxicity. The antioxidant N-acetyl-l-cysteine somewhat reduced 4-HPR cytotoxicity but did not affect ceramide species increase. We conclude the following: (a) 4-HPR was active against ESFT cell lines in vitro at concentrations achievable clinically, but activity was decreased in hypoxia; and (b) combining 4-HPR with ceramide modulators synergized 4-HPR cytotoxicity in normoxia and hypoxia.