Exploring the components of bleeding outcomes in transfusion trials for patients with hematologic malignancy
Clinically significant bleeding in patients with hematologic malignancies is a heterogeneous composite outcome currently defined as World Health Organization (WHO) bleeding Grades 2, 3, and 4. However, the clinical significance of some minor bleeds categorized as WHO Grades 1 and 2 remains controversial. We analyzed the number and frequency of individual signs and symptoms of WHO Grades 1 and 2 bleeds and explored their association with more severe incident bleeds graded as WHO Grades 3 and 4. STUDY DESIGN AND METHODS We aggregated daily bleeding assessment data from three randomized controlled trials conducted in patients with hematologic malignancies that used bleeding as an outcome. Cox proportional hazard regression analysis was used to identify signs and symptoms categorized as WHO Grades 1 and 2 bleeds that were associated with more severe bleeds (Grades 3 and 4). RESULTS We collected data from 315 patients (n = 5476 daily bleeding assessments; 3383 [61.8%] with a bleed documented). A total of 98.3% (3326/3383) were Grade 1 and 2 bleeds and 1.7% (57/3383) were Grades 3 and 4. Grade 1 and 2 bleeds were composed of 20 different bleeding signs and symptoms. Hematuria (hazard ratio, 16.1; 95% confidence interval, 4.4-59.2; P < .0001) was associated with incident Grade 3 or 4 bleeds. CONCLUSION In patients with hematologic malignancy, only hematuria (microscopic and/or macroscopic) was associated with more severe incident bleeds. This findings require validation in independent data sets.
Perioperative oral eltrombopag versus intravenous immunoglobulin in patients with immune thrombocytopenia: a non-inferiority, multicentre, randomised trial
Patients with haematologic malignancies and chemotherapy-induced thrombocytopenia (3 randomised controlled trials (RCTs), n= 315).
Less severe bleeds (WHO Grades 1 and 2).
More severe bleeds (WHO Grades 3 and 4).
The total data aggregated from the 3 RCTs was 5,476 daily bleeding assessments, 61.8% with a bleed documented. A total of 98.3% were Grade 1 and 2 bleeds and 1.7% were Grades 3 and 4. Grade 1 and 2 bleeds were composed of 20 different bleeding signs and symptoms. Haematuria was associated with incident Grade 3 or 4 bleeds. In patients with haematologic malignancy, only haematuria was associated with more severe bleeds.
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.
A randomized controlled trial comparing standard- and low-dose strategies for transfusion of platelets (SToP) to patients with thrombocytopenia
A noninferiority study was performed comparing low-dose and standard-dose prophylactic platelet transfusions. A double-blind randomized controlled trial (RCT) was performed in 6 sites in 3 countries. Thrombocytopenic adults requiring prophylactic platelet transfusion were randomly allocated to standard-dose (300-600 x 10(9) platelets/product) or low-dose (150- < 300 x 10(9) platelets/product) platelets. The primary outcome (World Health Organization [WHO] bleeding > or = grade 2) was assessed daily through clinical examination, patient interview, and chart review. A WHO grade was assigned through adjudication. The Data Safety Monitoring Board stopped the study because the difference in the grade 4 bleeding reached the prespecified threshold of 5%. At this time, 129 patients had been randomized and 119 patients were included in the analysis (58 low dose; 61 standard dose). Three patients in the low-dose arm (5. 2%) had grade 4 bleeds compared with none in the standard-dose arm. WHO bleeding grade 2 or higher was 49. 2% (30/61) in the standard-dose arm and 51. 7% (30/58) in the low-dose group (relative risk [RR], 1. 052; 95% confidence interval [CI], 0. 737-1. 502). A higher rate of grade 4 bleeding in patients receiving low-dose prophylactic platelet transfusions resulted in this RCT being stopped. Whether this finding was due to chance or represents a real difference requires further investigation. These clinical studies are registered on (http://www. clinicaltrials. gov) as NCT00420914.
A multicenter pilot-randomized controlled trial of the feasibility of an augmented red blood cell transfusion strategy for patients treated with induction chemotherapy for acute leukemia or stem cell transplantation
BACKGROUND Anemia may be an important factor contributing to an increased risk of bleeding, particularly in patients with thrombocytopenia. STUDY DESIGN AND METHODS A multicenter, single-blinded pilot randomized controlled trial (RCT) was performed to evaluate the feasibility of conducting a larger RCT to determine the effect of the hemoglobin (Hb) concentration on bleeding risk. Patients with acute leukemia receiving induction chemotherapy or those undergoing stem cell transplantation were assigned to one of two treatment groups: standard transfusion strategy (transfusion of 2 units of red blood cells [RBCs] when their Hb level was less than 80 g/L) or an augmented transfusion strategy (transfusion of 2 units of RBCs when their Hb level was less than 120 g/L). RESULTS Sixty patients were enrolled: 29 in the control group and 31 in the experimental group. The proportions of patients experiencing clinically significant bleeding and the time to first bleed were not significantly different between the control and experimental groups. The experimental group received more RBC transfusions (transfusions/patient-day) than the control group (0. 233 vs. 0. 151; relative risk, 1. 56; 95% confidence interval, 1. 16-2. 10; p = 0. 003). The proportion of patient-days with platelet (PLT) transfusions was not different between the experimental and control groups. The mean number of donor exposures (PLT and RBC transfusions) was not different between experimental and control groups. Bleeding symptoms were systematically documented. CONCLUSION This pilot study thus indicated that it would be feasible to enroll the required number of patients to enable the performance of a large RCT to investigate the effect of Hb on bleeding risk in thrombocytopenic patients.
Comparing the efficacy and safety of apheresis and whole blood-derived platelet transfusions: a systematic review
BACKGROUND A systematic review and meta-analysis was performed to determine if there were differences between apheresis platelet concentrates (APCs) or platelets (PLTs) derived from whole blood (WBD) for the outcomes acute reactions, alloimmunization, refractoriness, corrected count increment (CCI), radiolabeled recovery and survival, time to next transfusion, and bleeding. STUDY DESIGN AND METHODS We searched Medline, Embase, the Cochrane Registry of Controlled Trials, PapersFirst, ProceedingsFirst, and AABB and ASH abstracts for randomized controlled trials (RCTs) comparing APCs and WBD PLTs for clinical outcomes. Study selection, data extraction, and methodologic quality assessments were performed in duplicate. Results were pooled using meta-analytic methods. RESULTS Ten RCTs met the inclusion criteria. Acute reactions per patient were lower for APCs (relative risk [RR], 0.65; 95% CI, 0.44-0.98); however, when controlling for leukoreduction, there was no significant difference (leukoreduced [LR]-APCs vs. LR-WBDs; odds ratio, 1.78; 95% CI, 0.87-3.62). There was no difference between products when reaction frequencies were assessed per transfusion (RR, 0.65; 95% CI, 0.33-1.28). APCs were associated with significantly higher CCIs than WBD PLTs at both 1 hour (weighted mean difference [WMD], 2.49; 95% CI, 2.21-2.77) and 18 to 24 hours (WMD, 1.64; 95% CI, 0.60-2.67). No conclusions could be made for the outcomes of alloimmunization and refractoriness. No studies addressed outcomes of time to next transfusion or bleeding. CONCLUSIONS Owing to the small number of trials and lack of comparability of PLT products for leukoreduction, we were unable to draw definitive conclusions about the clinical benefits of APCs compared with WBD PLTs. Rigorous RCTs using clinically important end points are needed to settle this issue.
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.