AIM: To investigate the effects of 2-(8-hydroxy-6-methoxy-1-oxo-1H-2-benzopyran-3-yl) propionic acid (NM-3) alone and in combination with carboplatin on tumor growth and apoptosis in mouse models of human gastric cancer constructed by subcutaneous implantation of histologically intact tumor tissue.METHODS: Human gastric cancer SGC-7901 tissues were implanted into the dcrsal subcutis of nude mice.One week after tumors reached to a volume of 50-100 mm3 for around 1 wk, these mice were randomly divided into 8 groups (n = 10). NM-3 was injected peritoneally at the dose of 10 mg/kg, 20 mg/kg or 40 mg/kg every other day for 5 wk, combined with carboplatin (5 mg/kg) every third day for 4 wk. As controls of combined treatment,another 4 groups of mice were injected with either NM-3at 10 mg/kg, 20 mg/kg or 40 mg/kg, or with carboplatin alone (5 mg/kg). The control mice received normal saline. Tumor weight, tumor growth inhibition (TGI), and intratumoral microvessel density (MVD) were evaluated.Apoptosis of human gastric cancer was detected by TUNEL method and flow cytometry analysis, respectively.RESULTS: The mean tumor volume (692.40 ± 58.43 mm3, 548.30 ± 66.02 mm3, 382.13 ± 43.52 mm3)after treatment with carboplatin combined NM-3 at the dose of 10 mg/kg, 20 mg/kg or 40 mg/kg was lower than that after treatment with either NM-3 at the dose of 10 mg/kg, 20 mg/kg or 40 mg/kg or with carboplatin alone. Compared with the normal saline group, NM-3 administered at 10 mg/kg, 20 mg/kg or 40 mg/kg significantly reduced the tumor weight in these groups (P ＜ 0.05). Carboplatin used alone at 5 mg/kg showed minimal effects. But NM-3 in combination with carboplatin had greater effects of tumor weight than either NM-3 or carboplatin alone. NM-3 alone at the dose 10 mg/kg or in combination with carboplatin had no obvious effects on body changes. Two mice died of diarrhea in each of the two groups treated with 40 mg/kg NM-3 or with 40 mg/kg NM-3 in combination with carboplatin. A significant increase in apoptosis was observed in the NM-3 treated groups, and the effect was more significant in the groups treated with carboplatin in combination with NM-3 at 10 mg/kg, 20 mg/kg and 40 mg/kg, than in the control group. The induction of apoptosis was positively associated with the dose of NM-3. NM-3 significantly reduced the neomicrovascular formation of gastric cancer. The MVD was lower in the groups treated with NM-3 or with NM-3 in combination with carboplatin than in the group treated with carboplatin or in the normal saline group (P ＜ 0.05).CONCLUSION: The results suggest that the inhibitory effect of NM-3 on gastric cancer growth is mediated through decreased angiogenesis and the increased induction of apoptosis. Furthermore, NM-3 alone at the dose of 10 mg/kg or in combination with carboplatin has no obvious effects on body changes, indicating that NM-3 in combination with carboplatin may be effective in the treatment of gastric cancer. The toxicity of NM-3 needs further studies.