Search

Clinical trials of immunotoxins in cancer patients have been limited in many cases by vascular leak syndrome (VLS). Recently, rats were identified as a model for VLS induced by BR96 sFv-PE40, a carcinoma-reactive single-chain immunotoxin. In this study, the toxin component of this immunotoxin, PE40, was found to be responsible for inducing hydrothorax in rats, thereby demonstrating that direct binding to the BR96 antigen was not essential to the onset of VLS. Mutational analysis of PE40 determined that both ADP ribosylation and proteolytic processing functions innate to Pseudomonas exotoxin A (PE) were necessary for PE40 to induce hydrothorax in rats; however, neither function by itself was sufficient for VLS induction. Additionally, nonsteroidal anti-inflammatory agents were found to block VLS in rats receiving BR96 sFv-PE40. These results demonstrate that the toxin component of PE-based immunotoxins induce VLS and suggest agents for clinical management of the toxicity.

Immunotoxins are hybrid molecules composed of a cell-surface binding domain and a protein toxin moiety that together target specific cell populations for elimination. These agents represent a promising approach for the treatment of many human diseases, most notably cancer. However, it has recently become clear that many immunotoxins when used in human clinical trials induce vascular leak syndrome (VLS), restricting the administration of doses necessary to achieve good therapeutic responses. The lack of an appropriate animal model has hindered efforts to understand and prevent immunotoxin-induced VLS. We have found that in rats, intravenous administration of the single-chain immunotoxin BR96 sFv-PE40 results in symptoms that closely resemble VLS seen in human immunotoxin trials. A large fluid accumulation in the thoracic cavity was observed, along with an increase in hematocrit and body weight and a decrease in serum albumin. The VLS was apparent within 24 hr after administration of immunotoxin and was seen in both immunocompetent and athymic rats. Similar symptoms were not found in mice even at lethal doses. Prophylactic administration of the corticosteroid dexamethasone resulted in prevention of VLS and survival of rats injected with what would otherwise be lethal doses of BR96 sFv-PE40. Prophylactic treatment with dexamethasone in rats xenografted with human tumors either did not inhibit or minimally inhibited the antitumor activity of BR96 sFv-PE40. The use of prophylactic corticosteroids should be considered for immunotoxin clinical trials, since it may improve therapeutic efficacy by decreasing the dose-limiting toxicity of VLS.