Platelets and eltrombopag: a not-so-sticky situation

In this issue of Blood, Psaila and colleagues examine the in vivo effects of eltrombopag, a thrombopoietin receptor agonist (TPO-RA), on platelet function and reactivity.¹ 

Immune thrombocytopenia (ITP) is an autoimmune disease associated with increased platelet destruction and ineffective thrombopoiesis that presents with variable amounts of bleeding. Thrompobopoietin (TPO) is a hematopoietic growth factor that stimulates megakaryopoiesis and platelet production by activating the cell through the TPO-receptor (TPO-R, c-MPL). In 2008, two small molecule TPO-R agonists (TPO-RAs), eltrombopag and romiplostim, were approved for use in adult patients with chronic (> 12 months) ITP and low platelet counts. TPO-R is present not only on hematopoietic stem cells and megakaryocytes but also on platelets, and stimulation of platelets with TPO potentiates platelet activation to numerous agonists.²  Clinical trials examining the efficacy and safety of both eltrombopag and romiplostim have reported adverse events of thrombosis at a rate of 2% to 4% although the overall rates of thrombosis are similar between control and treated patients.³  Several case reports have also been published describing significant thrombosis in patients treated with TPO-RAs, prompting concerns about the potential for platelet activation by TPO-RAs.

Eltrombopag is a small molecule, oral TPO-RA that interacts with the transmembrane domain of TPO-R leading to JAK/STAT pathway signaling in platelets and megakaryocytes.⁴  A previous in vitro study suggested that eltrombopag, in contrast to TPO, did not alter reactivity of platelets⁴  and the authors suggested that a study examining the in vivo effects of eltrombopag is needed to better understand the potential for platelet activation in the target population.⁴  Psaila et al have answered the call and examined the in vivo effect of eltrombopag on platelets.¹ 

Studies of platelet function in patients with thrombocytopenia are difficult as low platelet counts may significantly affect the results of standard platelet aggregometry studies. Many studies have examined the utility of flow cytometry to interrogate platelet function.⁵  Psaila et al use flow cytometry to study the reactivity of platelets even at very low platelet counts⁶  and compare flow cytometry results in both control subjects with normal platelet aggregation studies and in thrombocytopenic patients with ITP. In control subjects, there was a direct correlation between platelet aggregation by light transmission aggregometry and flow cytometric platelet function. They used glycoprotein (GP) Ib (CD42) surface expression, which is probably most correlated with mean platelet volume/surface area of the platelets and is expected to decrease due to proteolysis on platelet activation, activated GPIIb/IIIa (CD41) by PAC-1 binding,¹  and P-selectin expression, a marker of platelet α-granule release,⁷  to study platelet function.

Psaila et al show that patients with untreated ITP have platelets that are more activated by flow cytometry than control patients as evidenced by increased expression of P-selectin and activated GPIIb/IIIa. There also was increased expression of GPIb, but this may reflect the larger mean platelet size noted. While the authors did not compare this platelet reactivity with patients who were thrombocytopenic for other reasons, these findings are consistent with prior reports that there may be an increased risk of thrombosis associated with ITP,⁸  that many patients with ITP do not have significant bleeding even with severe thrombocytopenia,⁹  and that platelets in patients with ITP may be activated.¹⁰  Interestingly, Psaila et al found that stimulation of platelets from ITP patients with the thrombin receptor activator TRAP resulted in lower responsiveness as measured by surface levels of activated GPIIb/IIIa and P-selectin expression, suggesting that while the platelets of ITP patients are more activated in the circulation, the potential for further activation of these platelets is limited in thrombocytopenic patients with ITP.

Next, the authors examined the effect of eltrombopag on platelet reactivity compared with control platelets and to the ITP platelets before treatment. First, treatment with eltrombopag did not change expression of GPIb or activated GPIIb/IIa expression on platelets in the absence of stimulation. There was a small increase in P-selectin early after treatment with eltrombopag (by day 7). Over time, platelet reactivity by activated GPIIb/IIIa expression to stimulation by ADP was decreased. Reactivity to TRAP measured by activated GPIIb/IIIa expression and P-selectin expression was increased with treatment in patients whose platelet count responded to eltrombopag compared with baseline, but did not increase above control levels.

These data do not support eltrombopag increasing platelet reactivity in ITP and may ameliorate some of the reactivity otherwise seen. This first in vivo report on TPO-RAs on platelet reactivity only studied 20 patients; however, these data are consistent with previously published studies of the in vitro effect of eltrombopag on platelet activation⁴  and with another, even smaller study recently published examining 3 patients that also demonstrated no obvious platelet activation with treatment with eltrombopag.¹¹  Nevertheless, based on the limited size of the present study, a further study of a larger cohort of patients to examine both the effect of ITP on platelet function as well as modification of this effect by the new TPO-RA is needed.