Kirk and Harlan reply

Kawai and colleagues report a considerable number of thromboembolic events in monkeys after administration of a mouse monoclonal antibody against human CD154. This group's extensive experience with the monkey kidney allograft model, coupled with their unprecedented run of thrombotic complications, does suggest that something is awry, perhaps that the antibody preparation used in their experiments is prothrombotic. Indeed, a link between CD154:CD40 and coagulation makes great teleological sense. Nevertheless, what makes sense and what is fact are often very different. However appealing, given the reported presence of CD154 on platelets and endothelia, we believe it premature to definitively conclude that the results discussed in this letter are epitope-specific.

Thromboembolic complications have been found after the initial use of many antibodies in humans and monkeys, and have largely disappeared with the use of refined antibody purification processes. Biogen's early clinical trials have been marred by thromboembolic events, indicating that humanized antibody against human CD154 induces thrombosis. These complications have been publicly disclosed and the trials have been halted pending additional pre-clinical evaluation. However, trials (admittedly different in design) using an antibody against human CD154 produced by IDEC Pharmaceuticals have proceeded so far without reported thromboembolic events.

Agents interfering with the CD154: CD40 pathway obviously powerfully influence the immune system and will no doubt have collateral effects. Potential effects, such as the ones presented here, must be scrutinized methodically and then corroborated by rigorous prospective investigation. Only then can important clues derived from anecdotes be accepted as fact.