A multicenter, randomized, controlled trial was conducted comparing active treatment with placebo in 227 patients with breast, colorectal, lung and other solid forms of cancer. Combination therapy, (CT) conventionally employed for the various types of tumor involved, was associated with MPA (117 patients) or placebo (110 patients). MPA was given orally as tablets, as a dose of 500 mg b.i.d. for 6 months. The results were, briefly, as follows: The incidence of leukopenia was significantly lower in the groups receiving MPA in patients with breast and colorectal cancer (P less than 0.02). Tumors of the lung and other solid forms showed no such difference. The incidence of thrombocytopenia was the same in all disease groups. Objective responses (CR + PR) were observed in 23/46 (50%) of breast cancer patients treated with CT + MPA, and in 13/47 (28%) of those given CT + placebo. The difference was significant (P less than 0.02). Subjective parameters also showed more improvement in the MPA group than in the patients given CT alone. No significant differences were found for the other types of tumor, but the numbers in this population were very limited. In a group of 45 patients, antithrombin III a (%), antithrombin III R: Ag (%), plasminogen (mg/dl), alpha-2 macroglobulin (%), factor VIII C (%), factor VIII R: Ag (%) and factor IX C (%) were determined. The most interesting post-treatment findings were an increase in anti-thrombin III (activity and antigen level) and in plasminogen. This suggests that MPA does not increase the risk of thrombosis, and might even, to some extent, impede tumor-induced thrombophilia.(ABSTRACT TRUNCATED AT 250 WORDS)