To my knowledge, no prognostic factors predict successful or poor outcomes after accessory splenectomy. I would suspect that patients with large splenic remnants or who have late relapses after splenectomy would be more likely to respond to accessory splenectomy, but these suspicions are not based on any data. I would also suspect that the response rate of 57%, as calculated by Dr. Kao, is high because reports of successful outcomes to treatment are more likely to be published than those of unsuccessful ones. In any case, I believe that accessory spleens should be removed (unless there are medical contraindications) because the treatment of refractory patients with other forms of therapy is associated with considerable morbidity; successful accessory splenectomy obviates this morbidity.

As stated by Dr. Pavletic, the use of allogeneic or autologous transplantation may offer another treatment approach for carefully selected patients with severe refractory chronic immune thrombocytopenic purpura. As a physician who spends most of his time involved with autologous and allogeneic transplantation, I would add a note of caution. Patients with this condition who are considered candidates for transplantation should be chosen very carefully because these procedures are associated with their own morbidity and mortality (probably <5% with autologous transplantation and >10% with allogeneic transplantation). In addition, should the antiplatelet autoantibody persist in a patient after transplantation, the platelet count may not recover quickly; thus, platelet support would be difficult or impossible. Clearly, a small group of patients with chronic immune thrombocytopenic purpura is resistant to all other forms of therapy, has almost constant mucosal bleeding, and will almost certainly die of the disease. This is the group that should be considered for transplantation. Unfortunately, these patients are also the ones who are most likely to have transplantation-related complications and those who are least likely to respond.