Perioperative oral eltrombopag versus intravenous immunoglobulin in patients with immune thrombocytopenia: a non-inferiority, multicentre, randomised trial
The Lancet. Haematology. 2020;7(9):e640-e648
BACKGROUND Patients with immune thrombocytopenia are at risk of bleeding during surgery, and intravenous immunoglobulin is commonly used to increase the platelet count. We aimed to establish whether perioperative eltrombopag was non-inferior to intravenous immunoglobulin. METHODS We did a randomised, open-label trial in eight academic hospitals in Canada. Patients were aged at least 18 years, with primary or secondary immune thrombocytopenia and platelet counts less than 100 × 10(9) cells per L before major surgery or less than 50 × 10(9) cells per L before minor surgery. Previous intravenous immunoglobulin within 2 weeks or thrombopoietin receptor agonists within 4 weeks before randomisation were not permitted. Patients were randomly assigned to receive oral daily eltrombopag 50 mg from 21 days preoperatively to postoperative day 7 or intravenous immunoglobulin 1 g/kg or 2 g/kg 7 days before surgery. Eltrombopag dose adjustments were allowed weekly based on platelet counts. The randomisation sequence was generated by a computerised random number generator, concealed and stratified by centre and surgery type (major or minor). The central study statistician was masked to treatment allocation. The primary outcome was achievement of perioperative platelet count targets (90 × 10(9) cells per L before major surgery or 45 × 10(9) cells per L before minor surgery) without rescue treatment. We did intention-to-treat and per-protocol analyses using an absolute non-inferiority margin of -10%. This trial is registered with ClinicalTrials.gov, NCT01621204. FINDINGS Between June 5, 2013, and March 7, 2019, 92 patients with immune thrombocytopenia were screened, of whom 74 (80%) were randomly assigned: 38 to eltrombopag and 36 to intravenous immunoglobulin. Median follow-up was 50 days (IQR 49-55). By intention-to-treat analysis, perioperative platelet targets were achieved for 30 (79%) of 38 patients assigned to eltrombopag and 22 (61%) of 36 patients assigned to intravenous immunoglobulin (absolute risk difference 17·8%, one-sided lower limit of the 95% CI 0·4%; p(non-inferiority)=0·005). In the per-protocol analysis, perioperative platelet targets were achieved for 29 (78%) of 37 patients in the eltrombopag group and 20 (63%) of 32 in the intravenous immunoglobulin group (absolute risk difference 15·9%, one-sided lower limit of the 95% CI -2·1%; p(non-inferiority)=0·009). Two serious adverse events occurred in the eltrombopag group: one treatment-related pulmonary embolism and one vertigo. Five serious adverse events occurred in the intravenous immunoglobulin group (atrial fibrillation, pancreatitis, vulvar pain, chest tube malfunction and conversion to open splenectomy); all were related to complications of surgery. No treatment-related deaths occurred. INTERPRETATION Eltrombopag is an effective alternative to intravenous immunoglobulin for perioperative treatment of immune thrombocytopenia. However, treatment with eltrombopag might increase risk of thrombosis. The decision to choose one treatment over the other will depend on patient preference, resource limitations, cost, and individual risk profiles. FUNDING GlaxoSmithKline and Novartis.
Assessing the effectiveness of whole blood-derived platelets stored as a pool: a randomized block noninferiority trial
BACKGROUND Prestorage pooling of whole blood-derived platelets (PLTs) would simplify bacterial detection. This study evaluated the in vivo effect of the prestorage pooling of PLTs stored for up to 5 days, by assessing the corrected count increment (CCI) 18 to 24 hours after transfusion of the product. STUDY DESIGN AND METHODS A randomized block noninferiority design was used. Eligible patients had chemotherapy-induced thrombocytopenia and were considered likely to need at least six PLT transfusions. For every block of two transfusion events, one consisted of PLTs stored individually and then pooled before transfusion, and the other was a product pooled before storage. The primary outcome was categorized as a successful (>4. 5) or unsuccessful ( RESULTS Twenty-three eligible patients received a total of 189 PLT transfusions. The median number of PLT transfusions was 7 (range, 0-27). Eighty-five complete transfusion pairs were used in the primary analysis. The proportions of transfusions leading to a CCI of greater than 4. 5 was identical for both routine and PLTs pooled before storage (45/85=52. 9%; relative risk, 1. 00; lower limit of the one-sided 95% confidence interval [CI], 0. 83). The estimate of the mean difference in CCI between pooled and routine storage (pooled-routine) was -0. 45 (95% CI, -2. 23 to 1. 33; p=0. 63). CONCLUSION These results provide evidence that storage of PLTs as a pool for up to 5 days results in posttransfusion CCIs that are not inferior to PLTs stored individually.