No abstract

SummaryClinicians must promptly decide which patients suspected of having heparininduced thrombocytopenia (HIT) warrant a change in anticoagulation. This single-centre series of 246 HIT testing referrals assessed the combination of clinical score (thrombocytopenia, timing, thrombosis, other causes of thrombocytopenia not evident; 4T's), Diamed ID-Heparin-PF4 immunoassay (PaGIA) and 14C Serotonin Release Assay (SRA) to develop a practical and safe diagnostic strategy for HIT. A total of 142/256 (58%) referrals were in patients with a low 4T's score, with 12/246 (5%) in the high scoring group. PaGIA was positive in 24/246 (9AE7%) patients, whilst SRA was positive in 9/246 (3AE6%). The overall positive predictive value of a positive PaGIA test alone was 37AE5%, however this reached 80% for the high scoring group. Both negative PaGIA and low clinical score independently had negative predictive values of 100%. We subsequently developed an algorithm that, when applied to this cohort, would have resulted in 18/246 patients (7%) definitely requiring alternative anticoagulation, whilst a further 7/246 (2AE8%) patients would have been considered on an individual basis. Ultimately (based on SRA) this would have resulted in 16/246 (6AE5%) patients unnecessarily having a change in their anticoagulation, with 9/246 (3AE6%) patients being 'correctly treated'. The combination of 4T's scoring and PaGIA permitted a practical and safe approach to rapid HIT diagnosis and management.Keywords: Heparin, heparin-induced thrombocytopenia (HIT), heparin platelet factor 4 antibody (HPF4), 4 T's clinical score, particle gel immunoassay. research paper ª

We have conducted a series of multilaboratory surveys during the last 6 years to evaluate testing proficiency in the detection of congenital and acquired thrombophilia. For lupus anticoagulant (LA) testing, participant laboratories used a panel of tests, including activated partial thromboplastin time (aPTT; 100% of laboratories), kaolin clotting time (26 to 70%), and Russell's viper venom time (RVVT; 75 to 100%). Coefficients of variation (CVs) for assays ranged from 5 to 40%. RVVT assays appeared to be most sensitive and specific for detection of LA (fewer false-negatives or -positives), although laboratories performed best when they used a panel of tests. For congenital thrombophilia, tests evaluated comprised protein C (PC), protein S (PS), antithrombin (AT), and activated protein C resistance (APCR). Most participant laboratories performed PC using chromogenic (approximately 75%), or clot based (approximately 15%) assays, with few (< 10%) performing antigenic assessments. PS was most often assessed (approximately 60%) by immunological or antigenic assays, usually of free PS, or by functional or clot-based assays (approximately 40%). AT is usually assessed by functional chromogenic assays (approximately 95%). APCR was assessed using aPTT (approximately 50%) or RVVT (approximately 50%) clot-based assays, with the aPTT APCR typically performed using factor V-deficient plasma predilution, but the RVVT APCR typically performed without. Laboratories using the RVVT APCR generally performed better in detection of factor V Leiden-associated APCR, with the aPTT method group yielding higher false-negative and/or false-positive findings (approximately 5% of occasions). Some clot-based PC and PS assays appeared to be influenced by APCR status, and yielded lower apparent PC and PS levels with positive APC resistance. The overall error rate for PC, PS, and AT was approximately 2 to 8% (i.e., false-normal interpretations for deficient plasma or false-abnormal interpretations for normal plasma). The CVs for these assays ranged from 5 to 40%, with highest CVs typically obtained with PS assays.

An improved activated factor X-based clotting method was used to investigate activity of procoagulant phospholipid (PPL) in blood samples collected into various anticoagulants and in plasmas with a range of abnormalities. The dilute activated factor X-activated clotting time (XACT) was carried out on a mixture of specimen with phospholipid-free porcine plasma. PPL from the test sample is then rate-limiting, controlling the clotting time so that the XACT is shortened from a maximum of approximately 120 s with citrated platelet-free plasma to approximately 30 s with freeze-thawed platelet-rich plasma. XACT results were only shortened slightly by fresh normal platelet-rich plasma, but were shortened significantly by platelets that had been activated by freeze-thawing. This improved method for PPL was not prolonged by deficiencies of known clotting factors and therapeutic levels of heparin, and it was surprisingly resistant to most lupus anticoagulants. However, it was extremely sensitive to PPL, detecting down to 50 ng/ml synthetic phospholipid added to phospholipid-free plasma. Excellent correlation was achieved between XACT shortening and microparticle count assessed by annexin V-binding and flow cytometry in normal plasma spiked with platelet microparticles. In citrated blood specimens, XACT shortened with time in a temperature-dependent manner. XACT results on blood samples anticoagulated with ethylenediamine tetra-acetate were more stable, and these would be preferable for assessing PPL expression ex vivo. XACT was significantly shorter in whole blood samples than in normal platelet-rich or platelet-poor plasmas, suggesting that PPL was normally expressed more by cells or aggregates larger than platelets or microparticles.