Association of Bleeding Scores and Platelet Transfusions With Platelet Counts and Closure Times in Response to Adenosine Diphosphate (CT-ADPs) Among Preterm Neonates With Thrombocytopenia

Platelet transfusions (PTX) are frequently given to neonates with thrombocytopenia to prevent bleeding.1,2 However, there is a poor association between platelet counts (PCs) and bleeding in neonates,2-4 suggesting that other factors are more important for bleeding risk than PCs. We previously showed5 that closure times in response to adenosine diphosphate (CT-ADPs), an in vitro test of primary hemostasis, were associated with concurrent bleeding scores (BSs) in preterm neonates, while PCs were not.5 In the current study, we analyzed the association of CT-ADP with BS over time and the associations of PTX with changes in PC, CT-ADP, and BS in neonates younger than 27 weeks’ gestational age.


Introduction
Platelet transfusions (PTX) are frequently given to neonates with thrombocytopenia to prevent bleeding. 1,2 However, there is a poor association between platelet counts (PCs) and bleeding in neonates, [2][3][4] suggesting that other factors are more important for bleeding risk than PCs.
We previously showed 5 that closure times in response to adenosine diphosphate (CT-ADPs), an in vitro test of primary hemostasis, were associated with concurrent bleeding scores (BSs) in preterm neonates, while PCs were not. 5 In the current study, we analyzed the association of CT-ADP with BS over time and the associations of PTX with changes in PC, CT-ADP, and BS in neonates younger than 27 weeks' gestational age.  [3][4][5][6][7][8][9][10][11][12][13][14] days) were enrolled in the Neonatal Hemorrhagic Risk Assessment in Thrombocytopenia Study. Patients were eligible if they were less than 32 weeks' gestation or weighed less than 1500 g at birth, had a PC less than 100 × 10 3 /μL (to convert to ×10 9 /L, multiply by 1.0), and had written informed consent from a parent or guardian.

Methods
Patients were excluded if they were not expected to survive, were thought to have a congenital thrombocytopenia or platelet dysfunction, or had a major chromosomal anomaly. Eligible neonates were enrolled between May 2015 and September 2017, 5 and PC, CT-ADP, and BS were measured on 3 consecutive days in each infant. Bleeding was quantified with the Neonatal Bleeding Assessment Tool, providing a BS ranging from 0 (no bleeding) to 4 (major bleeding). Details of all methods have been published previously. 5 Ordinal multinomial logistic regression was used to quantify the association of BS (grouped 0-1, 2, and 3-4) with the prior day's CT-ADP. To assess the association of 1-day changes in BS, CT-ADP, and PC with incident transfusion, we used linear regression and correlation. All analyses were conducted in SAS version 9.4 (SAS Institute), and we used repeated-measures models to account for withinpatient covariance. Statistical significance was set at P < .05, and all tests were 2-tailed.

Results
Baseline characteristics of the Neonatal Hemorrhagic Risk Assessment in Thrombocytopenia cohort have been reported previously. 5  a longer CT-ADP was associated with a higher probability of grade 2 to grade 4 bleeding the following day, rising from 10% after minimum CT-ADP (ie, 0 seconds) to 40% after maximum CT-ADP (ie, 300 seconds). Each 60-second increment in CT-ADP was associated with greater odds of higher BS the next day (odds ratio, 1.42; 95% CI, 1.01-2.00; P = .04), but adjustment for PC reduced the estimate of effect size (odds ratio, 1.33; 95% CI, 0.96-1.87; P = .09).
Changes in CT-ADP were strongly correlated with changes in BS (r = 0.33; P = .008), while changes in PC were not (r = -0.01; P = .93) (Figure). In regression analyses, changes in CT-ADP were   (Table).

Discussion
The main finding of this study was an association between changes in BS and changes in CT-ADP (but not PC), suggesting that primary hemostasis and bleeding are dynamic and more interconnected than PC and bleeding in extremely preterm neonates with thrombocytopenia. Limitations of our study included the small number of infants with PCs less than 50 × 10 3 /μL because of platelet transfusion practices in our neonatal intensive care units and the relatively small number of the more clinically relevant bleeding grades (ie, [3][4]. The limited number of CT-ADPs obtained in each infant was because of the relatively high blood volume required by the test (800 μL). We also recognize that all associations presented may be subject to residual or unmeasured confounding variables.
Platelet transfusions given at the thresholds used in this study increased PCs but did not reduce BSs. This was consistent with the Platelet Transfusion Thresholds in Premature Neonates (PlaNeT-2) study, in which PTX for PCs less 50 × 10 3 /μL did not reduce bleeding. 6 Together, these findings suggest that when PCs are not extremely low, other factors could be contributors to bleeding in neonates with thrombocytopenia, although the PC threshold below which a PTX reduces bleeding is unknown. Overall, 8 infants in our cohort also received 9 fresh frozen plasma transfusions for clinical bleeding. However, these transfusions did not reduce CT-ADP or BS and did not affect their association. Implementing CT-ADP or other measures of primary hemostasis in clinical practice may lead to novel approaches to manage thrombocytopenia in preterm neonates.