We thank Fakhouri and colleagues for commenting on our article and sharing their experience with six cases of TTP treated with plasma infusion. They question the validity of Rock and coworkers' study, which compared a plasma infusion regimen with a plasma exchange regimen, both in the first cycle and over long-term follow-up (1). We agree that the patients receiving plasma exchange did receive larger volumes of plasma, but the duration of therapy in the first cycle was similar and the outcomes were different. At the end of the first treatment cycle, the response rate (defined by an increase in platelet count) was higher in patients who received plasma exchange (24 of 51 patients) than in those who received plasma infusion (13 of 51 patients) (P 0.025). Two of the 51 patients treated with plasma exchange died, compared with 8 of the 51 patients who received plasma infusion (P = 0.035).

Rock and colleagues' study was a valid assessment of which of the two then-current approaches (plasma exchange or plasma infusion) was superior in the treatment of TTP. As noted in that report, the patients in the plasma exchange group did receive greater volumes of plasma because of the limits of the volume of plasma that could be tolerated in those who received infusion. As we alluded to in our report, the presence of antibodies to von Willebrand factor-cleaving protease and the role of von Willebrand factor multimers in adhesion and subsequent microthrombosis provide a mechanism for the superiority of plasma exchange over plasma infusion. Plasma exchange can remove the antibodies to the protease as well as the von Willebrand factor multimers and would be expected to have greater therapeutic benefit than plasma infusion, which can provide only von Willebrand factor multimer-cleaving protease (2, 3).

We are pleased that the correspondents' patients with TTP did well after the first cycle of plasma infusion. However, we would be remiss if we did not mention that it is difficult to accurately predict which patients presenting with TTP are at risk for death. We also point out that in a study of 108 patients with TTP and the hemolytic-uremic syndrome, most deaths occurred soon after diagnosis (4).

Thus, although plasma infusion may be effective in some patients, the difficulty of accurately identifying the patients who are at highest risk for death has led us to use plasma exchange as a standard form of therapy.