A Canadian colleague reports that an attending hematologist/oncologist at her hospital has begun to order "trial" platelet transfusions for patients with a presumptive diagnosis of ITP who are scheduled for surgery. This has happened so far on two separate occasions for two separate patients. The 'justification' is "to see if the patient will respond if platelets were needed during a planned surgery". According to the Canadian colleague, both patients had platelet counts in the 80,000/uL to 90,000/uL range, and when she queried the hematologist/oncologist before the first "trial" platelet dose, he became irate to be questioned. She laments that she just gave out the second "trial" platelet transfusion without discussing it with the ordering physician since her lab director did not back her up in the first case. In neither case did the trial platelet transfusions cause sustained increments of the platelet counts. Reportedly, the platelet count did not go up in the first case (and actually dropped to 50,000/uL after 5 days). The second case did not show an increase in platelet count in the first day, and went down to about 60,000/uL on the third day. The inquiring colleague wonders if there are new data or practice pathways/guidelines to support the strategy of using a trial platelet transfusion for patients with a presumptive diagnosis of ITP.

The following comments have been received.

ADDENDA April 21, 2007

1. A colleague in Vienna, Austria reports that in his opinion a patient with ITP is very unlikely to respond with a significant corrected count increment if transfused with platelets. Furthermore, a trial dose of platelets for a patient who has had transfusions or been pregnant in the past may boost previously stimulated antibodies (such as HLA antibodies) which can become problematic if they cause the patient to be refractory to platelet transfusions. Consequently, he thinks that a trial platelet transfusion to see how a patient with ITP will respond is not indicated, and may be harmful.

2. Dr. Neil Blumberg of the University of Rochester School of Medicine (attribution used with permission) reports that he is unaware of any data to support the practice of transfusing a 'trial' dose of platelets to patients with a presumptive diagnosis of ITP. In fact, he comments that "there is very little evidence base of any sort for transfusion therapy in ITP, and that's being charitable". That having been said, from his own 30 years or so of experience with ITP, he thinks there are very few situations where the risk of a platelet transfusion is necessary to assess clinical response prior to surgery, or necessary at all in ITP. He adds that if a patient with ITP has a platelet count greater than 50,000 per µL, the number of patients with ITP with counts in this range who have CLINICALLY abnormal hemostasis is small. Anecdotally, in the days when bleeding times were available and reasonably reliable, he reports having seen patients with ITP with counts in the 5,000-20,000/µl range who had normal bleeding times, and underwent major surgery with no platelet transfusions, and no excess bleeding. He vividly recalls a patient he saw with Dr. Peter McPhedran as a resident who had a C-section with a platelet count of 5,000/µl, no transfusions, and no bleeding before or after the operation. Not necessarily typical, but instructive. Unfortunately, the current substitute in vitro tests for platelet function, while convenient and reproducible, (e.g., PFA100) are not useful in such settings due to technical limitations. Thus, in his experience, the single best predictive assessment of surgical hemostasis (in general, not only in ITP) is simply the patient's bleeding history and current physical exam. If there is no evidence of compromised hemostasis (no purpura, petechia---particularly in the mucous membranes of the mouth), the likelihood of significant bleeding at surgery is very low indeed. There is simply no substitute for a history and physical exam in assessing hemorrhagic potential prior to invasive procedures. He concludes saying that platelet transfusion is probably only indicated in ITP for life threatening hemorrhage or emergency invasive procedures/surgery where hemostatic function clearly isn't normal by clinical criteria.

ADDENDA April 25, 2007

3. Dr. Maurice Genereux, Medical Director of Cangene Corporation reports that the company has not undertaken any clinical studies nor seen any published data to support the administration of a "trial" of platelet transfusion for patients with a presumptive diagnosis of ITP scheduled for surgery. Dr. Genereux reports that in his experience the role of platelet transfusions for ITP is very disappointing and should perhaps be limited to the treatment of catastrophic hemorrhages with other therapies such as corticosteroids and Immune Globulins. The algorithm presented by Douglas Cines and Victor Blanchette in a Review Article on Immune Thrombocytopenic Purpura: N Engl J Med, Vol. 346, No. 13. March 28, 2002. makes a similar recommendation.

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator

W. Tait Stevens, MD
CBBS e-Network Forum Assistant Editor & Moderator