Effects of Anti-Glycoprotein Antibodies on Response of Immune Thrombocytopenia Patients to Thrombopoietin Receptor Agonists and on Megakaryocytes Viability

Background

Immune Thrombocytopenia (ITP) is a bleeding disorder due to a combination of increased platelet destruction and reduced production, often secondary to anti-platelet/megakaryocyte antibodies.

The presence of antibodies to glycoproteins (GP) IIb/IIIa (integrin α_IIbβ₃) and GPIb/IX, detected in majority of ITP patients, may correspond to different responses to treatment, i.e., anti-GPIb is associated with more severe disease, and less responsive to intravenous immunoglobulins and steroids. 

Thrombopoietin Receptor Agonists (TPO-RA) increase platelet production by stimulation of megakaryopoesis. Predictors of response to TPO-RA and influence of antibody profile on response are currently unknown. In our previous study we investigated Absolute Immature Platelet Fraction (A-IPF) prior to TPO-RA treatment and did not find a correlation between A-IPF, anti-GP antibodies, and platelet counts.

The aims of this study were to further investigate:

1.      The role of anti-GP antibodies in response to TPO-RA;

2.      Effect of patients' antibodies on megakaryocyte (MK) viability, maturation, apoptosis  and formation of proplatelets (in vitro);

3.      The influence of patients' clinical characteristics on response to TPO-RA.

Materials and Methods

91 patients with persistent or chronic ITP, were treated at Weill Medical College of Cornell University until January 2015 with TPO-RAs: 52 patients received eltrombopag, 22 romiplostim and 17  avatrombopag . Serum samples of 84 patients were analyzed for the presence of anti-GP by MAIPA assay as previously described.

Patients with baseline platelet counts less than <30x10⁹/L were defined as responders to TPO-RA if the average of their six median monthly platelet counts was ≥50x10⁹/L and doubled from average baseline counts (prior to TPO-RA). Patients with baseline platelet counts 30-50x10⁹/L were responders if the average platelet count was ≥75x10⁹/L.

MKs were derived from human umbilical cord blood stem cells as previously described.  Cells were grown using SFEM medium, adding on day 0 of culture 50 ng/ml recombinant TPO and aliquots of serum of ITP patients or healthy controls. The percentages of immature (CD41+/CD42-), mature (CD41+/CD42+), viable and apoptotic MKs were analyzed by flow cytometry on day 12. Apoptosis was analyzed by measuring Mitochondrial Outer Membane Potential (MOMP) and Phosphatidyl Serine (PS) externalization. MKs were considered apoptotic if they had positive staining for PS externalization, viable if positive for MOMP, and dead if positive for 7-Aminoactinomycin D (7AAD).

Proplatelet formation by MKs was analyzed by microscopy.

Statistical analysis using unpaired T-test and Pearson correlation test were performed.

Results

 

Ninety-one patients were included, 40 male (44%) and 51 female (56%), with a median age of 37.4 years (range 2-87). Median duration of ITP before TPO-RA treatment was 8 years (range 0.3-45).  

The 18/91 (19.8%) non-responders to TPO-RA were not different from the 73/91 responders in age, gender, number of prior treatments, duration of ITP, and past splenectomy. The presence of either or both anti-GP antibodies was correlated with average lower platelet counts on TPO-RA: 82.3 x10⁹/L versus 123x10⁹/L in patients without detected antibodies ("neither") (p=0.003). However, the response to TPO-RA was not influenced by the type of antibody: in patients with anti-GPIb the average platelet count was 76.1x10⁹/L, and with anti-GPIIb/IIIa 80.7x10⁹/L (Figure 1). In culture, excess dead MKs were found in anti-GPIb group and antiGPIb&antiGPIIb/IIIa group compared to "neither" group (p=0.0013 and p=0.027 respectively) and comparing  antiGPIb&antiGPIIb/IIIa to control (p=0.0025). We did not observe changes in the degree of MK apoptisis or in MK maturation in the presence of serum antibodies. In cultures treated with serum of patients having anti-GPIb, less proplatelets were detected comparing to control (p=0.044) or to "neither" (p=0.0039).

We conclude that patients with anti-GP antibodies respond less to TPO-RA, however there is no difference in response to TPO-RA between patients having anti-GPIb and anti-GPIIb/IIIa, unlike responses to other treatment modalities (e.g., steroids or immunoglobulins). TPO-RA could be a preferable treatment option in ITP patients having anti-GPIb.

Figure 1. Average 6-months platelet counts of TPO-RA-treated ITP patients divided into groups by presence of antibody/ies.