When preparing patients with immune thrombocytopenia purpura (ITP) for invasive procedures, there are a few key considerations. First, while such patients may be asymptomatic with chronic thrombocytopenia, they typically require a higher platelet count than their baseline in advance of their procedure. The optimal platelet target is uncertain; however, platelet counts higher than 50 × 109/L (lower-risk bleeding procedure) or higher than 80-100 × 109/L (higher-risk) are often cited as target thresholds.1 Second, the amount of time available to prepare for interventions may have implications for what agents are used, depending on expected onset of action. The time to response of intravenous immunoglobulin (IVIG) is approximately 24 to 48 hours, while corticosteroids may take three to four days in those who respond.2 Third, accessibility of agents such as IVIG in resource-constrained environments may be challenging. Finally, these agents may have toxicities that must be accounted for (e.g., IVIG may cause headache, allergic reactions, and aseptic meningitis; steroids may impact adversely on hyperglycemia, wound healing, and mood disturbance).3
For these reasons, oral eltrombopag was recently studied in the Bridging ITP Trial as an alternative for perioperative optimization. This thrombopoietin (TPO) receptor agonist has been studied in chronic ITP and has a typical onset of action of seven to 14 days.4 This was a multicenter, open-label, noninferiority randomized controlled trial at eight Canadian academic hospitals. Patients with chronic primary or secondary ITP undergoing minor or major surgery were randomly assigned to receive either IVIG or eltrombopag before their procedure. IVIG was given on day –7 relative to procedure date (+/– 2 days) at a total dose of 1-2 g/kg; additional “rescue” doses were permitted before day –1 and up to day +7 after the procedure. Eltrombopag was initiated at 50 mg once daily on day –21 before surgery and administered until seven days after surgical hemostasis, with dose adjustments made based on weekly platelet counts. The primary outcome was achievement of target platelet count higher than 45 × 109/L (lower-risk bleeding procedure) or higher than 90 × 109/L (higher-risk bleeding procedure) from day –1 to day +7 without requiring rescue therapies (including corticosteroids, platelet transfusions, and IVIG).
Seventy-four patients were randomized to either eltrombopag (n = 38) or IVIG (n = 36). Approximately 85 percent of patients had primary ITP, with 8 to 21 percent requiring chronic prednisone use at baseline and baseline platelet counts of approximately 40 × 109/L in both groups. Thirty-nine to 45 percent of procedures were major surgery (including 25% overall undergoing splenectomy). The primary outcome by intention-to-treat analysis was reached in 30 (79%) of 38 patients (eltrombopag group) versus 22 of (61%) 36 patients (IVIG group), meeting the criteria for noninferiority (p=0.005). A post-hoc analysis demonstrated superiority of eltrombopag over IVIG (absolute risk difference, 17.8%; one-sided lower limit of 95% CI 0.4%; p=0.047) by intention-to-treat analysis.
Among the 38 patients receiving eltrombopag, the median time to platelet target was 12 days; 37 percent reached that target within seven days, and 53 percent reached it within two weeks. Meanwhile, among the 36 patients receiving IVIG, median time to platelet target was six days. Treatment failures occurred in eight patients receiving eltrombopag and in 14 patients receiving IVIG (10/14 at day 0, 4/14 between day 0-7). Rescue therapy was required in seven patients in each group, and severe bleeding events occurred in a similar portion of patients in each group (24% eltrombopag, 22% IVIG group). One patient in the eltrombopag group sustained a treatment-related pulmonary embolism 14 days after skin biopsy, during which platelet count was 71 × 109/L. One additional patient in the IVIG group developed deep vein thrombosis 30 days after hip arthroplasty while receiving mechanical thromboprophylaxis; this was adjudicated to be unrelated to IVIG, and platelet count was 81 × 109/L.
There were a few limitations with this study that have been noted in an accompanying editorial.5 First, given the short time to onset of IVIG, an initial dose given closer to the surgical date (3-4 days) may have been preferable and may have affected treatment failure rates. Second, the designations of periprocedural bleeding risk were assigned by local investigators, which may have led to inconsistency (laparoscopic splenectomy was listed as both a minor and major surgery). Finally, there was a slow recruitment phase.
These results suggest that eltrombopag may be a reasonable alternative for optimizing platelet counts before surgery, as long as there is adequate preparation time (3 weeks). These data are also in alignment with prior observational studies with romiplostim and eltrombopag that suggest they are efficacious and safe in this context.6,7 However, one must take thrombotic risk factors into account because one patient receiving eltrombopag did develop treatment-related pulmonary embolism. The possible thrombotic risk of TPO agonists has been noted in other studies as well.8 The use of TPO receptor agonists postoperatively after splenectomy with the risk of postoperative thrombocytosis should be cautiously observed.
In my view, in situations where there is limited time to prepare for surgery (< 1-2 weeks), the preferred option would be IVIG owing to its more rapid onset of action, tolerability, and safety profile. If there is more time (3+ weeks) available before elective surgery, eltrombopag could be considered if access to IVIG is limited and there are no additional high-risk thrombotic risk factors, given that the two options ultimately seem to have similar overall rates of rescue therapy and postoperative bleeding. The authors must be congratulated for completing the first randomized trial for perioperative ITP management.
Dr. Tseng indicated no relevant conflicts of interest.