Discussion

The role of platelets in ductal closure has first been demonstrated in murine models.^{Reference Echtler, Stark and Lorenz5} Observational studies on the clinical significance of this finding in preterm infants have yielded conflicting results. However, two recent meta-analyses indicate a moderate association between low platelet counts in preterm infants and failure of spontaneous and pharmacologically induced ductus arteriosus closure.^{Reference Simon, van Zogchel, Bas-Suarez, Cavallaro, Clyman and Villamor7,Reference Mitra, Chan and Paes8} We previously demonstrated in 1430 very low birth weight infants that low platelet counts within the first 24 hours after birth were not associated with patent ductus arteriosus, while low platelet counts before and during cyclooxygenase inhibitor treatment were moderately associated with pharmacological ductus arteriosus treatment failure.^{Reference Sallmon, Weber and Huning9,Reference Sallmon, Weber and Dirks10} Interestingly, a recent randomised controlled trial could not demonstrate any positive effect of platelet transfusions on patent ductus arteriosus in preterm infants.^{Reference Kumar, Dutta, Sundaram, Saini, Sharma and Varma16} We and others therefore hypothesised that impaired platelet function, rather than platelet number, contributes to spontaneous or pharmacologically induced ductus arteriosus closure in preterm infants.^{Reference Sallmon, Weber, von Gise, Koehne and Hansmann17–Reference Sallmon, Barikbin, Koehne, von Gise and Hansmann19} Two recent studies support this hypothesis: Collagen-adenosine-diphosphate closure times > 130 seconds (platelet function analyser) were independently associated with ductal patency.^{Reference Kahvecioglu, Erdeve and Akduman20} Moreover, impaired platelet function due to inflammation may contribute to failure of ductal closure in preterm infants with infection.^{Reference Meinarde, Hillman, Rizzotti, Basquiera, Tabares and Cuestas21}

The immature platelet fraction is a modern laboratory parameter that assesses the number of newly released platelets from peripheral blood samples. Immature platelet fraction measurements have been reported in preterm neonates and may prove useful in guiding platelet transfusion therapy.^{Reference Cremer, Paetzold and Schmalisch11,Reference Cremer, Weimann, Schmalisch, Hammer, Buhrer and Dame12,Reference MacQueen, Christensen and Henry22} Similar to the reticulocyte count in anaemia, the immature platelet fraction increases in response to thrombocytopenia, if megakaryopoiesis and thrombopoiesis are intact.^{Reference Cremer, Sallmon, Kling, Buhrer and Dame6} A recent study found an association between immature platelet fraction (percentage) and bleeding risk in children with immune thrombocytopenic purpura, independently from platelet counts.^{Reference McDonnell, Bride, Lim, Paessler, Witmer and Lambert23} The latter results support previous findings on immature platelets having distinct functional properties which affect their haemostatic potential.^{Reference Lador, Leshem-Lev, Spectre, Abelow, Kornowski and Lev13}

Based on these observations, we hypothesised that, due to their distinct functional properties, mature and immature platelets are independently associated with incidence of haemodynamically significant ductus arteriosus in preterm infants. Here, for the first time, we show that lower absolute counts of mature platelets on days 3 and 7, but not on day 1 after birth, are associated with ductal patency.^{Reference Sallmon, Weber and Huning9,Reference Sallmon, Weber and Dirks10}

In addition, we found higher immature platelet fraction (percentage) values in infants with haemodynamically significant patent ductus arteriosus as compared to those without. However, logistic regression analysis revealed that only mature platelet counts on postnatal day 7, in contrast to immature platelet fraction values, were independently associated with haemodynamically significant patent ductus arteriosus in our cohort. Our results suggest that lower counts of mature platelet contribute to a higher incidence of haemodynamically significant patent ductus arteriosus. In contrast, immature platelet fraction values, which can be regarded as a physiological response to thrombocytopenia in preterm infants, do not independently affect ductal closure. Of note, we found similar results indicating a role for lower mature platelet counts in failed spontaneous ductal closure when comparing all infants with patent ductus arteriosus (not only haemodynamically significant ones) to those without a patent ductus arteriosus (data not shown).

Our study is limited by its relatively low sample size. Thus, we were unable to assess a potential association between low mature platelet counts and treatment failure as only 24 infants underwent pharmacological treatment for patent ductus arteriosus in our cohort. In addition, the current study did not address a potential association between inflammation, as seen in septic preterm neonates, mature and immature platelets, and ductus arteriosus. Furthermore, ductal closure is a process that is characterised by a complex interplay between molecular and morphological changes within the vessel wall, which, in addition to platelets, is triggered by vascular adhesion and migration of mononuclear blood cells to the muscular layer of the vessel wall.^{Reference Waleh, Seidner and McCurnin24} Assessing those complex molecular and cellular processes requires a highly sophisticated experimental design and is beyond the scope of this paper. Thus, further research is required to elucidate the complex interplay between immature and mature platelets on ductal closure and to assess their specific functions and association with bleeding risk and thromboembolic complications in preterm infants.^{Reference Ibrahim, Schutt, Hannawi, DeLao, Barker and Kleiman25,Reference Matter, Ragab, Roushdy, Ahmed, Aly and Ismail26}