Immature platelet fraction as a pre‑test probability marker for anti‑platelet IgG positivity in thrombocytopenia

Abstract Background: The differential diagnosis of thrombocytopenia is a common clinical challenge. The immature platelet fraction (IPF), measured automatically in routine hematology analyzers, reflects thrombopoietic activity and is elevated in cases of immune-mediated platelet destruction such as immune thrombocytopenia (ITP)1. Anti-platelet IgG flow cytometry assays are highly specific (> 90%) but suffer from low sensitivity (~53%) and still have a debated role in thrombocytopenia investigation2. We investigated whether IPF could serve as a triage marker to guide test ordering in thrombocytopenic patients undergoing anti-platelet antibody testing. Methods: We retrospectively identified all adult patient samples from September 2021 to May 2025 with platelet counts ≤150 × 10^9/L who had anti-platelet IgG testing for investigation of thrombocytopenia at the clinical laboratory of Einstein Hospital Israelita. Patients investigated for congenital platelet disorders were excluded. IPF was quantified using Sysmex XN analyzers via the PLT-F fluorescent channel. Anti-platelet IgG was detected by direct flow cytometry: platelet-rich plasma incubated with anti-CD41 PE and anti-IgG FITC, analyzed on a DxFLEX cytometer (Beckman Coulter). Positivity was defined as a mean fluorescence intensity (MnX) greater than two standard deviations (SD) above the mean of five normal controls. Patients with doubtful results were excluded from the final analysis. Demographics were recorded. IPF distributions were compared with Mann–Whitney U; association with antibody positivity measured by point-biserial correlation and logistic regression (odds ratio per 1 % IPF). Discrimination was assessed via ROC/AUC and Youden index, with 95% CIs via bootstrap (1 000 iterations), and decision-curve analysis assessed net benefit. Subgroups stratified by platelet ranges (<30, 30–50, 50–100, 100–150 × 10^9/L). Results: We included data from 376 patients (249 positive, 127 negative); median age 45 (IQR 30–62), 58% female, with no significant demographic differences between groups. Median IPF was significantly higher in the antibody-positive group (10.0% vs. 7.25%; p=0.0002), though the correlation was weak (r=0.187). The overall ROC AUC was 0.614 (95% CI 0.553–0.674), with an optimal threshold of ~7.6% yielded sensitivity 0.659, specificity 0.535, positive predictive value (PPV) 0.778, and negative predictive value (NPV) 0.381. The odds ratio was ≈1.06 per 1% increase in IPF. In the severe thrombocytopenia subgroup (<30 × 10⁹/L; n=68), the AUC improved to 0.726; a cutoff of ~18.7% provided 100% specificity but only 48% sensitivity. Other subgroups showed lower discriminatory performance. (AUC 0.504–0.606). Decision-curve analysis indicated a modest net benefit for IPF-guided triage only at high pre-test probabilities (>0.6). Discussion: In this cohort of thrombocytopenic patients, IPF is significantly elevated in those with positive anti-platelet IgG antibodies, confirming its association with an immune-mediated peripheral platelet destruction. While this association is statistically significant on a population level, the modest diagnostic accuracy (AUC = 0.623) indicates that IPF has limited utility as a standalone test to predict or exclude the presence of anti-platelet antibodies for an individual patient. However, its effectiveness is enhanced in patients with severe thrombocytopenia (<30 × 10⁹/L), where the AUC increased to 0.726. Within this group, a high IPF cutoff of ~18.7% demonstrated high specificity (1.00), providing a strong indicator for an underlying immune process. Technical limitations of IPF include its dependency on analyzer platform and known variability in precision at extremely low platelet counts. For the antibody assay, limited sensitivity, potential interference from plasma proteins, and lack of standardization between labs remain concerns. We conclude that IPF is modestly predictive of anti-platelet IgG positivity and may serve as contextual pre-test probability tool in thrombocytopenic patients, with its greatest utility found in patients with severe platelet reductions. To our knowledge, this is one of the first studies to detail this association across different platelet strata. Prospective validation and integration with clinical variables into multivariable diagnostic pathways are warranted. 1- doi:10.1111/j.1365-2141.2004.04987.x 2- doi: 10.1111/jth.14419