A whole blood model of thrombocytopenia that controls platelet count and hematocrit

Abstract

In patients with thrombocytopenia, it can be difficult to predict a patient's bleeding risk based on platelet count alone. Platelet reactivity may provide additional information; however, current clinical assays cannot reliably assess platelet function in the setting of thrombocytopenia. New methods to study platelet reactivity in thrombocytopenic samples are needed. In this study, we sought to develop a laboratory model of thrombocytopenia using blood from healthy subjects that preserves the whole blood environment and reproducibly produces samples with a specific platelet count and hematocrit. We compared the activation state of unstimulated and agonist-stimulated platelets in thrombocytopenic samples derived from this method with normocytic controls. Whole blood was diluted with autologous red blood cell concentrate and platelet-poor plasma, which were obtained via centrifugation, in specific ratios to attain a final sample with a predetermined platelet count and hematocrit. P-selectin exposure and GPIIbIIIa activation in unstimulated platelets and platelets stimulated with collagen-related peptide (CRP) or adenosine diphosphate (ADP) in thrombocytopenic samples and the normocytic control from which they were derived were quantified by flow cytometry. Our methodology reliably produced thrombocytopenic samples with a platelet count ≤50,000/μL and an accurately and precisely controlled hematocrit. P-selectin exposure and GPIIbIIIa activation on unstimulated platelets or on ADP- or CRP-stimulated platelets did not differ in thrombocytopenic samples compared to normocytic controls. We describe a new method for creating thrombocytopenic blood that can be used to better understand the contributions of platelet number and function to hemostasis.

Keywords: Blood platelets; Flow cytometry; Platelet activation; Platelet aggregation; Thrombocytopenia.

Fig. 1

Dilution method of creating thrombocytopenic whole blood. From the starting whole blood, a small aliquot is set aside. The remainder of the whole blood is centrifuged at 300g for 8 min to separate the platelet-rich plasma (PRP) from the red blood cells (RBCs). The PRP is subsequently centrifuged at 3800g for 10 min, and the platelet-poor plasma (PPP) is removed from the platelet pellet. The PPP, RBCs, and previously set aside whole blood aliquot are mixed together in a ratio based on goal platelet count and hematocrit

Fig. 2

Hematocrit accuracy based on goal platelet count. Hematocrit is not affected by platelet count in the thrombocytopenic samples with platelet counts of 20,000/µL (black symbols), 50,000/µL (gray), and 100,000/µL (white). Data are expressed as mean ± standard deviation

Fig. 3

Effect of dilution versus centrifugation method for preparation of thrombocytopenic whole blood on the activation state of unstimulated platelets. P-selectin exposure on resting platelets was quantified in normocytic control samples (black circle) and thrombocytopenic samples in samples prepared using the dilution method (black square) or the centrifugation method (black triangle) using flow cytometry. Thrombocytopenic samples prepared using the centrifugation method had a higher levels of P-selectin exposure on platelets as indicated by median fluorescence intensity (MFI) and b a higher percent of platelets that were positive for P-selectin compared to both the normocytic control and platelets in thrombocytopenic samples prepared using the dilution method. The platelet activation state did not differ significantly in normocytic control samples and thrombocytopenic samples prepared using the dilution method. *P ≤ 0.01, **P ≤ 0.001, and ***P < 0.0001

Fig. 4

Whole blood aggregometry in thrombocytopenic samples before and after ex vivo transfusion. Area under the curve (AUC) in response to stimulation with ADP was significantly decreased in samples with a platelet count of 20,000 or 100,000/µL compared to normocytic controls. There was no significant difference in the AUC following ADP stimulation between thrombocytopenic samples to which platelets were added back to achieve a platelet count similar to the normocytic control and the unmanipulated normocytic control. Lines (gray) indicated normal range as supplied by the multiplate manufacturer. ***P ≤ 0.001, ****P < 0.0001, and ns not significant when compared to control

Fig. 5

P-selectin exposure and GPIIbIIIa activation in normocytic controls (black circle) and thrombocytopenic (black square) samples prepared using the dilution method There was no significant difference in a P-selectin exposure or b GPIIbIIIa activation in (from left to right) unstimulated platelets nor in platelets stimulated with CRP, low-(1.25 µM) and high-dose (5 µM) ADP, or samples pre-treated with apyrase prior to addition of high-dose ADP