Revisiting Platelet-Associated Tissue Factor: Methodological Concerns in Flow Cytometry Detection

Letter to: Tissue Factor, a Membrane-associated Marker of Platelet Activation, Predicts 5-year Cardiovascular Mortality in Coronary Artery Disease Patients Thromb Haemost eFirst DOI: 10.1055/a-2771-2148 10.1055/a-2771-2148 We read with great interest the paper by Camera et al., “ Tissue Factor, a Membrane-associated Marker of Platelet Activation, Predicts 5-year Cardiovascular Mortality in Coronary Artery Disease Patients. ”[ 1 ] The study addresses a critical unmet need in coronary artery disease: improving thrombotic risk stratification by incorporating dynamic biomarkers like platelet-associated tissue factor (TF). While the findings are clinically significant, they touch on a highly debated topic in hematology: the presence and origin of TF on platelets. The topic of TF expression on platelets is highly debated. While some researchers argue that TF-positive platelets represent a subset of platelets circulating in blood, containing intracellular TF protein that may be released by ADP-activation[ 2 ] and pre-mRNA/RNA,[ 3 ] others attribute the detection of TF on platelets to contamination by monocyte-derived extracellular vesicles.[ 4 ] [ 5 ] In this study, Camera et al.[ 1 ] utilized an in-house flow cytometry protocol[ 6 ] to detect TF on ADP-stimulated platelets. Their method employs the HTF-1 mouse monoclonal antibody against TF, followed by an Alexa Fluor (AF)633-conjugated goat anti-mouse secondary antibody. To ensure platelet specificity, the authors co-stained with mouse anti-CD41 PE and mouse anti-CD14 PerCP to exclude monocytes. While detailed, this approach raises a significant technical concern: the use of a polyclonal secondary anti-mouse antibody in a staining cocktail containing multiple mouse-derived primary antibodies risks generating false-positive signals. Specifically, the secondary antibody can bind to all mouse IgGs present in the cocktail (including anti-CD41 and anti-CD14), rather than specifically detecting the anti-TF primary antibody. To investigate this concern, we replicated the authors' protocol. Whole blood was collected in 0.129 M sodium citrate, aliquoted, and either stimulated with adenosine 5′-diphosphate (ADP, 200 μM) or left unstimulated. Samples were processed in the presence or absence of the anti-TF antibody (clone HTF-1). Following 15 minutes of incubation, samples were fixed with 1% paraformaldehyde for 1 hour at room temperature, washed, and stained with the secondary goat anti-mouse AF633 antibody and CD41-PE. As we wished to systematically control this analysis, additional controls were included: unstained cells, single-stained samples, and samples in which the primary anti-TF antibody was omitted or replaced with an unlabeled, irrelevant IgG isotype control, whereas the secondary goat anti-mouse AF633 antibody was still used. In all cases, sample preparation was identical, and platelet activation was confirmed using P-selectin staining. As shown in [ Fig. 1A ], single staining with anti-CD41 PE resulted in clear platelet detection within the characteristic FSC/SSC gate, which increases following ADP stimulation (histogram, [ Fig. 1A ]). It is important to note here that no compensation was applied in [ Fig. 1A ], as no fluorescence spillover from the PE channel was detected. Staining with the secondary antibody alone or following either an irrelevant IgG control or the anti-TF antibody, resulted in negligible background in the AF633 channel ([ Fig. 1B ], upper, middle, and lower panels, respectively). Importantly, no differences were observed between unstimulated and ADP-stimulated samples. However, as depicted in [ Fig. 1C ], a strong AF633 signal was detected whenever anti-CD41 PE was present, regardless of the presence of the primary anti-TF antibody. Furthermore, as shown in the histograms in [ Fig. 1C ], activation increased the CD41-PE signal, which spilled into the AF633 channel, and could be misinterpreted in these settings as higher levels of TF on platelets surface in ADP-activated blood. Fig. 1 Flow cytometry analysis of tissue factor expression in platelets using whole blood. ( A ) Gating strategy and detection of platelets in whole blood (WB) using single staining with anti-CD41 PE. The bottom histogram shows the increase of CD41 in platelets from ADP-stimulated blood. ( B ) A two-stage staining of control (unstimulated) and ADP-stimulated WB in which Alexa Fluor (AF)633-conjugated goat anti-mouse antibody is used alone (upper panel), or following preapplication of an irrelevant IgG control (middle panel), or anti-TF (HTF-1) primary antibody (lower panel). ( C ) A two-stage staining of control and ADP-stimulated WB with anti-CD41 PE and AF633-conjugated goat anti-mouse antibody used alone (upper panel), following preapplication of an irrelevant IgG control (middle panel), or anti-TF (HTF1) primary antibody (lower panel). The histograms further show the increase of the PE signal following activation of platelets in whole blood, which also affects the signal in the AF633 channel that is used for the secondary antibody. These findings suggest that the reported TF signal may result from the secondary antibody binding to the mouse anti-CD41 primary antibody, rather than true TF expression. While compensation can be used to subtract PE fluorescence from the AF633 channel, it cannot correct for the physical binding of a secondary antibody to the “wrong” primary antibody; doing so merely creates an analytical artifact. While the study by Camera et al. provides valuable insights into thrombotic risk stratification, our findings raise concerns about the methodology used to detect TF on platelets and the specificity of the signal. Given the clinical implications of using TF as a prognostic biomarker, it is essential to clarify whether the reported signal reflects true biological expression or a methodological artifact. Further validation of the protocol is warranted to ensure the robustness of these findings and their potential application in clinical practice. Publication History Received: 18 March 2026 Accepted: 31 March 2026 Article published online: 13 May 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany Comment to this article: Reply to the Comment: Revisiting Platelet-associated Tissue Factor: Methodological Concerns in Flow Cytometry Detection Thromb Haemost eFirst DOI: 10.1055/a-2862-6769