This study was conducted to determine whether reactivity to melanoma cells of pretreatment peripheral blood mononuclear cells (PBMCs) from patients with metastatic melanoma correlated with subsequent response to treatment with interleukin-2 (IL-2). The sensitivity of the quantitative real-time polymerase chain reaction (PCR) assay was optimized, including the total number of cells used (3 x 10(6) in 1 mL), the responder-to-stimulator cell ratio (5:1), the optimal time to incubate PBMCs with tumor (2 h), the appropriate tumor stimulators (melanoma cell lines differing only in the expression of histocompatibility leukocyte antigen [HLA-A2]), the duration of recovery in the culture of PBMCs after cryopreservation (18-24 h), and the medium used (Iscove, 10% human AB serum). Using this optimized assay to detect HLA-A2-restricted antitumor reactivity in the pretreatment PBMCs from patients with melanoma, positive reactive responses were detected in 7 of 28 patients with an objective clinical response to IL-2 therapy compared with 6 of 21 positive reactive responses in nonresponding patients. None of 12 healthy donors were positive in this study. Thus, there was no significant difference in the reactivity of pretreatment PBMCs when responders were compared with nonresponders, although the melanoma patients had an increased incidence of response compared with healthy donors (p = 0.05). The PBMCs from 11 of the 13 melanoma patients with pretreatment HLA-A2-restricted antimelanoma reactivity were tested against a panel of transfectants expressing known shared melanoma antigens. Anti-MART-1 reactivity was detected in the pretreatment PBMCs of three patients. It thus appears that some melanoma patients are immunologically primed to antigens expressed on the tumor surface, although the HLA-A2-restricted antimelanoma activity detected in this real-time PCR assay was not predictive of patients' responses to IL-2 therapy.