Fourty-eight patients entered the study, 37 of whom were evaluable, who underwent a nonrandomized combined modality approach plus a randomized immunotherapy treatment assigning the patients to one of the three arms: 1) thymostimulin, 2) beta IFN, 3) thymostimulin + beta IFN. The patients enrollment started in May 1991 and was closed in February 1993: the final evaluation was in February 1994. An immunological evaluation was made for all patients enrolled. The immunological assessment confirms some previous reports, such as the defective proliferative response of PBMC, the high serum level and the low production in culture of soluble IL 2 receptor (s IL 2 R), the low serum level and the low production of IL 2, and moreover shows new data. It is noteworthy that the significant increases--especially for cytokines and s IL 2 R, except for IL 1 alpha which has decreased, both for serum levels and for culture production--are induced by treatment with beta IFN (or with thymostimulin + beta IFN), suggesting that beta IFn is the immunologically effective moiety.