Twenty-seven patients with advanced gastrointestinal malignancies received recombinant gamma-interferon (rIFN-gamma, Biogen) prior to treatment with the murine monoclonal antibody 17-1A (Centocor), which mediates human monocyte antibody-dependent cellular cytotoxicity (ADCC). rIFN-gamma was used because it enhances human monocyte Fc receptor expression, nonspecific monocyte cytotoxicity (NSMC) and ADCC in vitro. The study was designed to identify a rIFN-gamma dose with acceptable toxicities which enhanced NSMC and ADCC. Patients received one course of therapy consisting of rIFN-gamma by 4-h infusions daily for 4 days at doses ranging from 0.001 to 80.0 X 10(6) units/m2/d, followed by 400 mg of 17-1A on day 5. The maximally tolerated dose of rIFN-gamma in this study was 40 X 10(6) units/d. Significant toxicity was seen at the high (greater than 1 X 10(6) units) but not low (less than or equal to 1 X 10(6) units) dose levels. Monocytes were isolated from patients' peripheral blood at baseline and on Days 3 and 5 for cytotoxicity studies which measured 111-In release from SW1116 cells which bear the target antigen of 17-1A. Low dose rIFN-gamma enhanced NSMC by Day 5 as well as did high dose therapy. ADCC enhancement was seen with low dose therapy (% specific lysis on Day 5 = 23.5 +/- 6.4 SEM versus baseline of 9.6 +/- 3.3, P = 0.03), but not with high dose rIFN-gamma treatment. Total (i.e., NSMC + ADCC) monocyte cytotoxicity was equivalent in the low and high dose treatment groups, although ADCC contributed more to total values in the low dose group. These findings were particularly striking if monocytes were exposed to additional rIFN-gamma in vitro prior to incubation with labeled target cells. We conclude that low dose rIFN-gamma therapy is at least equivalent, and possibly superior to high doses in this setting. Furthermore, low dose therapy, supplemented by ex vivo incubation of purified monocytes with rIFN-gamma, may be an optimal treatment strategy for this cytokine.