Previous results from our preclinical studies have shown that lonidamine (LND) can positively modulate the antiproliferative activity of doxorubicin (DOX) on breast cancer cell lines. To evaluate the effect of LND in a clinical setting, a multicenter randomized trial was carried out on patients with advanced breast cancer. From September 1991 to July 1993, 181 patients were enrolled in the trial and received an initial treatment of DOX at 75 mg/m2 for 3 cycles. The 137 patients who reached complete remission, partial remission, or stable disease were randomized to receive either DOX alone (75 mg/m2 day 1) (arm A) or DOX plus LND (600 mg orally/day) (arm B). The patients enrolled in the two arms were fairly homogeneous in terms of major clinical characteristics. Toxicity was similar in both arms except for myalgia: WHO grade > or=2 was observed in 57% of arm B patients. Overall response rate to DOX + LND was 50% and to DOX alone 38% in evaluable patients, and 48% vs 37% in all registered patients, as determined by an intention-to-treat analysis. The differences did not reach statistical significance. Conversely, in agreement with previous findings, we observed a significant difference in response rate in the subgroup of patients with liver metastases, regardless of the extent of hepatic involvement (DOX + LND 68% vs DOX 33%, p=0.03). This observation makes LND an important tool in association with anthracyclines in the treatment of this subgroup of patients.