Introduction: The differential diagnosis of hereditary and acquired thrombocytopenias can be challenging, especially when between immune thrombocytopenia (ITP) and less well characterized hereditary thrombocytopenias (HT) such as MYH9-related disorders. Fundamental differences in the management of these two conditions add clinical relevance to the search for novel parameters that differentiate these conditions. The immature platelet count (IPF) represents the fraction of platelets with higher RNA content, and in analogy to the reticulocyte count for erythropoiesis is a biomarker of thrombopoietic activity. In a recent report (Miyazaki et al, 2015), IPF values that were more than 5-fold higher than those observed in ITP patients were reported in a population of 15 patients with HT. However, whether this increased values represented a real increase in thrombopoietic activity, or reflected a technical limitation of IPF determination in large platelets could not be clarified. Here, we aimed to evaluate the role of IPF determination in the differential diagnosis between HT and several forms of acquired thrombocytopenia in a larger and more diverse population of patients. We also evaluated thrombopoietin (TPO) levels in HT compared to ITP, to further investigate the mechanisms by which extremely large IPF values are observed in HT.

Methods: IPF and mean platelet volume (MPV) were prospectively determined using a Sysmex XE5000 hematologic analyzer (as part of the complete blood count) in a cohort of patients with post-chemotherapy thrombocytopenia (n=56), bone marrow failure (myelodysplastic syndromes and aplastic anemia; n=22), ITP (ITP; n=105) and inherited thrombocytopenias (n=27). The latter population consists of a well-defined cohort of individuals with HT thrombocytopenia characterized by clinical, familial, laboratory and molecular data. TPO levels were determined by ELISA (R&D Systems) in 21 HT patients and 22 ITP patients matched for platelet count and age. A group of 178 healthy volunteers were used to determine normal IPF and MPV values.

Results: Median platelet counts were similar in post-chemotherapy patients (CTx) (32.0*109/L), bone marrow failure (BMF) (33.5*109/L), ITP (52.0*109/L) and HT (52.0*109/L) (P=0.15). Similar IPF levels were observed in CTx and BMF patients (5.6%; IQR 3.4-8.8% and 6.5%; IQR 3.5-13.7%. Compared to these two groups, higher IPF values were observed in both ITP (12.3%; 7.0-21.0%) and HT patients (29.8%; 17.5-56.4%) (both P values < 0.05). In addition, IPF were significantly higher in HT compared to ITP (Kruskall-Wallis test and Dunn's post test,P=0.001). MPV values were different between HT and CTx/BMF groups, but could not differentiate ITP from HT. TPO levels. The accuracy of IPF to discriminate HT from all other causes of thrombocytopenia estimated by ROC analysis was 0.88 (CI95%0.8-0.96, p<0.0001). Similar TPO levels were observed in platelet count-matched ITP, HT patients and healthy volunteers without thrombocytopenia. Interestingly, TPO presented marked correlations with both platelet count (Rs = - 0.61, P=0.002) and IPF (Rs= 0.59, P=0.003), even with TPO levels in the same range of healthy individuals. In contrast, no significant correlation could be observed between TPO and IPF or platelet count in HT patients.

Conclusions: IPF represents an informative biomarker for the differential diagnosis of hereditary and acquired thrombocytopenias, and accurately differentiates ITP from the most common HT. As expected, TPO levels in patients with ITP were not higher than in individuals with normal platelets counts. The inverse correlation between TPO and platelet count in these patients confirm a blunted TPO response to thrombocytopenia in these patients. Similarly, patients with HT did not present increased TPO levels compared to healthy individuals. Accordingly, increased IPF levels in these patients cannot be attributed to higher TPO levels.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.