Platelet production and platelet destruction: assessing mechanisms of treatment effect in immune thrombocytopenia

Abstract

This study investigated the immature platelet fraction (IPF) in assessing treatment effects in immune thrombocytopenia (ITP). IPF was measured on the Sysmex XE2100 autoanalyzer. The mean absolute-IPF (A-IPF) was lower for ITP patients than for healthy controls (3.2 vs 7.8 × 10⁹/L, P < .01), whereas IPF percentage was greater (29.2% vs 3.2%, P < .01). All 5 patients with a platelet response to Eltrombopag, a thrombopoietic agent, but none responding to an anti-FcγRIII antibody, had corresponding A-IPF responses. Seven of 7 patients responding to RhoD immuneglobulin (anti-D) and 6 of 8 responding to intravenous immunoglobulin (IVIG) did not have corresponding increases in A-IPF, but 2 with IVIG and 1 with IVIG anti-D did. This supports inhibition of platelet destruction as the primary mechanism of intravenous anti-D and IVIG, although IVIG may also enhance thrombopoiesis. Plasma glycocalicin, released during platelet destruction, normalized as glycocalicin index, was higher in ITP patients than controls (31.36 vs 1.75, P = .001). There was an inverse correlation between glycocalicin index and A-IPF in ITP patients (r² = -0.578, P = .015), demonstrating the relationship between platelet production and destruction. Nonresponders to thrombopoietic agents had increased megakaryocytes but not increased A-IPF, suggesting that antibodies blocked platelet release. In conclusion, A-IPF measures real-time thrombopoiesis, providing insight into mechanisms of treatment effect.

Figure 1

Immature platelet fraction (IPF) for ITP patients and controls. (A) Baseline IPF percentage in healthy controls (n = 100) and ITP patient treatment episodes (n = 29). The y-axis represents percentage IPF; and the x-axis, individual treatment episodes. Every pretreatment percentage IPF for all 29 patient episodes with ITP was greater than any of the 100 normal controls. Mean percentage IPF for ITP patients, 28.2% ± 15.5% was greater than controls, 3.2% ± 1.4% (P < .01). (B) Baseline A-IPF in healthy controls (n = 100) and ITP patient treatment episodes (n = 29). The y-axis represents the A-IPF values; and the x-axis, individual treatment episodes. Only 4 of 29 ITP patient episodes before treatment had A-IPF greater than 4 × 10⁹/L, to contrast to 93 of 100 controls. Mean A-IPF for ITP patients (n = 29), 2.8 ± 2.5 × 10⁹/L was less than the controls (n = 100), 7.8 ± 3.1 × 10⁹/L (P < .01).

Figure 2

Maximum observed change in the A-IPF within 10 days after treatment in patients with ITP (n = 24). The y-axis represents the maximum change in A-IPF after treatment, with an A-IPF response line threshold drawn at 10 × 10⁹/L; and the x-axis, those patients who had a platelet response to different treatments. All patients responding to Eltrombopag, none of the patients responding to GMA161, and 0 of 7 patients responding to intravenous anti-D had an increase in A-IPF more than the 10 × 10⁹/L threshold. Two of 8 patients treated with IVIG and 1 treated with IVIG and intravenous anti-D had an A-IPF increase of at least 10 × 10⁹/L.

Figure 3

ITP patients had similar glycocalicin levels but higher glycocalicin indicies (GCI) than controls with an inverse correction between GCI and A-IPF, and higher glycocalicin and A-IPF values for patients treated with TPO-A. (A) Plasma glycocalicin levels for patients with ITP compared with healthy controls. The y-axis represents plasma glycocalicin levels; and the x-axis, individual samples. There was no statistically significant difference between the mean (± SE) glycocalicin levels for ITP patients (n = 17) and controls (n = 8): 1.86 ± 0.25 versus 1.60 ± 0.21 μg/mL (P = .144). (B) GCIs for patients with ITP and healthy controls. The y-axis represents GCIs; and the x-axis, individual samples. There was a significant and large difference between the mean (± SE) GCI for ITP patients (n = 17) and controls (n = 8): 31.36 ± 13.28 versus 1.75 ± 0.24 (P = .001). (C) Correlative analysis of GCI and A-IPF for ITP patients. The y-axis represents GCI; and the x-axis, A-IPF. There was a negative correlation between GCI and A-IPF (r² = −0.578, P = .015). This demonstrates that platelet destruction is equivalent to platelet production. (D) Paired correlative analyses of plasma glycocalicin levels and A-IPF with platelet counts for ITP patients receiving thrombopoieitin receptor agonists and IVIG and/or prednisone. One line graph has a y-axis of A-IPF, and the other represents plasma glycocalicin levels, divided into those patients treated with TPO-A and those treated with IVIG and/or prednisone. The x-axis represents the platelet count. There were positive trends for A-IPF with platelet count for those patients treated with TPO-A (r² = 0.503, P = .216) and IVIG and/or prednisone (r² = 0.829, P = .058). This is shown in conjunction with negative trends for plasma glycocalicin levels with platelet count for those treated with TPO-A (r² = −0.611, P = .115) and for those treated with IVIG and/or prednisone (r² = −0.543, P = .297). Patients treated with TPO-A had greater A-IPF and plasma glycocalicin levels than those treated with IVIG and/or prednisone.

Figure 4

Intrapatient consistency for A-IPF and platelet count responses to anti-D and IVIG. (A) A-IPF and platelet count responses on 2 consecutive treatment episodes with intravenous anti-D in the same ITP patient. One line graph has a y-axis representing platelet count, and the other represents A-IPF, against an x-axis representing time. Dramatic increases in platelet counts with minimal corresponding increase in A-IPF, in response to anti-D treatment, are shown. This response was consistent for 2 consecutive treatment episodes in the same patient. (B) A-IPF and platelet count responses on 3 consecutive IVIG treatment episodes in the same ITP patient. One line graph has a y-axis representing platelet count, and the other represents A-IPF, against an x-axis representing time. The platelet count and A-IPF increased substantially after each IVIG administration. The platelet count and A-IPF responses to IVIG for this patient were consistent for all 3 treatment episodes.