To investigate the anti-tumor effect and the possible mechanism of arsenic trioxide (ATO) alone or in combination with thalidomide (THAL) for treatment of SCID mice model transplanted with human myelodysplastic syndrome (MDS) cell line MUTZ-1 cells.(1) The animal model was established in SCID mice; 75 SCID mice and 10 BALB/CA-nude mice were studied in this experiment. MUTZ-1 cells were cultured in vitro and made for mono-cell suspension (with 1 x 10(8)/ml cell density and in exponential growth behavior) and were subcutaneously implanted into 4-6-week-old first-generation SCID mice and BALB/CA-nude mice. The biological characteristics of the subcutaneous tumor cells were evaluated by the methods of cell morphology, histopathology, immunology by flow cytometer, chromosome analysis and immunohistochemistry (IHC). Subsequently, the tumor cells from first-generation mice model were respectively subcutaneously implanted into 61 second-generation SCID mice and 8 BALB/CA-nude mice and the rate of the tumor formation and the latent period of the tumor formation were observed. (2) In vivo, 56 MDS-SCID mice were randomly grouped; 40 of them were used as ATO treated groups [5.0 microg or 7.5 microg/(g.d) intraperitoneal injection (i p) 5 d a week, x 3 weeks] alone or in combination with THAL 8 microg/(g.d), x 3 week or THAL alone and 16 mice as the control groups [with 0.9% NaCl 10 microl/(g.d) i p or untreated]. The mean tumor diameters (MTD) of subcutaneous tumors were measured with slide gauge and the therapeutic effects and the survival period and the rates of survival were evaluated by the methods of histopathology, IHC, microvessel density count (MVD), DNA ladder, TUNEL and PI with flow cytometry.(1) The rate of the subcutaneous tumor formation was higher (98.4%, 60/61) in SCID mice than in BALB/CA-nude mice (62.5%, 5/8) (P = 0.0027). The latent period of the tumor formation was significantly longer (23 - 28 d, median 26 d) in BALB/CA-nude mice than that in SCID mice (10 - 17 d, median 12 d) (Z = 4.605, P0.001). The biological characteristics of the tumor cells in the MDS-SCID mice model were evaluated and considered as of anthropo-source and were consistent with that of MUTZ-1 cells, which showed that the MDS-SCID mice model was successfully established. (2) In vivo, the marked inhibitory effect on the subcutaneous tumors growth (F = 146.94, P = 0.000) and the higher rates of cells apoptosis were seen in the groups treated with ATO 5.0 microg or 7.5 microg alone or in combination with THAL than in control groups (F = 30.10, P = 0.000). The longer-survival periods (F = 25.11, P0.01) with lower toxicity were only observed in lower-dose group of ATO 5.0 microg than in other treated groups and control groups. THAL alone group had a mild inhibitory effect on the tumors growth (2 weeks) with a longer-survival period and higher-rate of survival than controls, but had no events of cell apoptosis. The expression of vascular endothelial growth factor (VEGF) protein and CD34 protein and MVD were markedly down-regulated by THAL compared with control groups (P0.01), suggesting that the possible mechanisms of the inhibitory effect on tumor growth by THAL related to inhibition of vascular endothelial growth. The legs paralysis in 2 MDS-SCID mice was observed after treatment with THAL alone (11 d and 15 d, respectively), which were considered as the side-effect of THAL associated with deep venous thrombotic (DVT) events and the mechanism for these events is unclear. The therapeutic effects were unsatisfied in the group treated with ATO 7.5 microg in combination with THAL due to more intense toxicity and the shorter-survival periods than other treated groups (P0.001).(1) The Hum-MDS-SCID mice model was successfully established, which served as an animal model for studying pathogenesis of MDS and therapeutic agents selection. (2) ATO had marked inhibitory effect on the subcutaneous tumors growth in MDS-SCID mice model in vivo. The longer-survival periods and higher-rates of survival with lower toxicity were observed in lower-dose ATO (5.0 microg) alone than other groups. The mechanisms of the anti-tumor effect of ATO were considered to be related to inducing cells apoptosis, but in the case of THAL, related to inhibition of vascular endothelial growth. (3) The results do not support the preconceived hypothesis of a synergistic effect of ATO (7.5 microg) in combination with THAL for treatment of MDS mice model due to the more intense toxicity and the shorter-survival periods in the treated group. Whether or not this will translate into clinically relevant effect of the ATO or THAL in MDS patients deserves further investigation.