Comparison of Five Thrombin Time Reagents

The thrombin time (TT) assay is a core test in clinical laboratories that perform coagulation testing. This assay screens for abnormalities in the conversion of fibrinogen to fibrin. The time necessary for fibrinogen to clot is affected by hypofibrinogenemia, dysfibrinogenemia, and the presence of inhibitors of the fibrinogen-to-fibrin reaction (heparin, hirudin, fibrin degradation products, and paraproteins) (1). The TT is used primarily to evaluate plasma specimens with prolonged activated partial thromboplastin time (APTT) values and, to a lesser extent, prolonged prothrombin time (PT) values for heparin or other thrombin inhibitors. The test is also useful to detect quantitative and qualitative fibrinogen abnormalities. Important criteria for selecting a TT reagent are high sensitivity to heparin, sensitivity to hypofibrinogenemia, and acceptable precision. We compared five commercial TT reagents on our automated coagulation analyzer and evaluated their sensitivity to fibrinogen and heparin concentration, their precision, and their accuracy. Information on each of the five TT reagents evaluated is provided in Table 1 of the data supplement (available with the online version of this Technical Brief at http://www.clinchem.org/content/vol49/issue1/). Each commercial reagent was reconstituted according to the manufacturer’s instructions and used to measure TT in singleton measurements on the STA-R coagulation analyzer (Diagnostica Stago) according to the manufacturer’s specifications and laboratory standards. Clotting times for the TT assay were initiated when the instrument pipetted 100 μL of patient plasma, which then incubated for 120 s. The addition of 100 μL of thrombin reagent activated the final measurement of analyte. The Diagnostica Stago and BioPool reagents required a longer incubation time of 240s. Fibrinogen was measured on the STA Compact (Diagnostica Stago) with use of the STA-Fibrinogen reagent (Diagnostica Stago) containing human thrombin of 100 000 IU/L (100 IU/mL). The assay was initiated when the plasma sample was diluted 1:20 (5 μL + 95 μL) with Owrens–Kohler buffer (0.028 mol/L sodium barbiturate–0.126 mol/L sodium chloride). The final measurement was taken after a 240-s incubation, with the addition of fibrinogen reagent (50 μL). Heparin was measured with an anti-Xa assay performed on the MDA (BioMerieux). This chromogenic assay was initiated by dilution (1:25; 2 μL + 48 μL) of the plasma sample with antithrombin III/buffer concentrate and incubation for 160 s. Factor Xa reagent (50 μL) was then added and incubated for 220 s. The instrument then added 50 μL of substrate, and the absorbance at 405 nm was measured. For each reagent, the following characteristics were evaluated: sensitivity to fibrinogen and heparin concentrations, precision (intra- and interrun), accuracy, and stability of reagents. The study was approved by the Institutional Review Board. Samples were deidentified. Blood was obtained by clean venipuncture. If an evacuated tube was used, then a pilot tube was drawn first. An exact ratio of 9 volumes of blood to 1 volume of anticoagulant (32 g/L citrate) was maintained. The anticoagulated specimen was then centrifuged at 3000g for 20 min. Samples were then aliquoted and kept frozen at −60 to −80 °C. Forty samples were collected in the same manner from healthy donors for reference interval studies. The sensitivity of each reagent to fibrinogen concentration was assessed by analyzing samples of pooled normal plasma diluted with fibrinogen-deficient plasma. Sample group 1 had fibrinogen of 570-3000 mg/L (57–300 mg/dL), whereas sample group 2 had fibrinogen of 390-2720 mg/L (39–272 mg/dL). After TT was measured on each sample, fibrinogen was measured. Our imprecision target was CV <10%. For heparin sensitivity, plasma samples from patients receiving heparin therapy were collected. To confirm the heparin concentration, heparin was also measured with an anti-Xa assay on the MDA (BioMerieux). Heparin concentrations were 50–600 IU/L (0.05–0.60 IU/mL). This study was performed over 3 days with three separate sets of patient samples. Our imprecision target was CV <10% for therapeutic heparin concentrations. Intrarun precision was evaluated with a normal and an abnormal control (n = 10). The interrun precision study was completed over a 5-day period. Each day, a normal and an abnormal control were analyzed in replicates of five. The stability of reagents was monitored over a 7-day period. On day 1, reagents were reconstituted and normal and abnormal controls were aliquoted into seven containers and frozen for each day’s use. Each day a normal and an abnormal control were thawed and analyzed, and the percentage of expected over observed was calculated. TT results were compared with an in-house method (Parke-Davis topical thrombin diluted with normal saline and glycerol; 30 000 IU/L). Clinical specimens that had previously been tested on the Fibrometer (Becton Dickinson) were tested concurrently with all five reagents on the STA-R analyzer. We assayed 35–40 samples encompassing the reportable range. Reference intervals were also estimated. We examined each reagent’s ability to detect plasma samples with low fibrinogen or with heparin. Measured fibrinogen concentrations were 570-3000 mg/L (57–300 mg/dL) on day 1 and 390-2720 mg/L (39–272 mg/dL) on day 2. Fibrinogen <1500 mg/L (150 mg/dL) is abnormal. TT results obtained with reagents from Diagnostica Stago, Sigma Diagnostic, and BioData all remained within the established reference interval until the fibrinogen was below this concentration. The BioMerieux and BioPool reagents produced slightly prolonged TT values for one of the two normal fibrinogen samples, for unclear reasons (Table 1 ). Intra- and interrun imprecision (CV) was <10% for fibrinogen measurements. Influence of fibrinogen and heparin on TT.1 For fibrinogen assays, separate samples were used on days 1 and 2. For heparin assays, plasma was obtained from patients receiving heparin therapy. A different patient sample was used on each of 3 days. Influence of fibrinogen and heparin on TT.1 For fibrinogen assays, separate samples were used on days 1 and 2. For heparin assays, plasma was obtained from patients receiving heparin therapy. A different patient sample was used on each of 3 days. We next evaluated heparin sensitivity, using plasma samples from patients receiving therapeutic unfractionated heparin. Our goal was to have a TT reagent capable of detecting heparin at 50 IU/L. The BioData thrombin reagent was highly sensitive to heparin. At heparin concentrations of 50 and 100 IU/L, the TTs were dramatically prolonged; at 200 IU/L heparin, the clotting time was >150 s. This reagent was considered too sensitive for monitoring heparin. The BioPool reagent was also quite sensitive to heparin, with TTs reaching 12.9 and 13.2 s at 50 and 40 IU/L, respectively. Although these clotting times are borderline normal and abnormal, respectively, with 100 IU/L heparin, TTs were increased and increased steadily with increasing heparin concentration. The Diagnostica Stago reagent was less sensitive to heparin than the BioData reagent. The Stago reagent produced a TT within the reference interval at 50 IU/L; results became abnormal at heparin concentrations >100 IU/L. The Sigma Diagnostics reagent was the least sensitive to heparin, producing clotting times within the reference interval at 50 and 100 IU/L; results became abnormal at 200 IU/L heparin. The BioMerieux reagent was more variable; on day 1, results obtained with that reagent became abnormal at 50 IU/L, and on day 2, results were within the reference interval until a heparin concentration of 120 IU/L was reached. The BioPool reagent was similarly variable (Table 1 ). The sensitivity of the TT reagents to heparin in the therapeutic range [300–600 IU/L (0.3–0.6 IU/mL)] was also evaluated. At a heparin concentration of 400 IU/L (0.4 IU/mL), both the BioData and Stago reagents produced TT values >150 s, whereas the BioMerieux, Sigma Diagnostics, and BioPool reagents produced TT values of 33, 30, and 59 s, respectively. At a heparin concentration of 550 IU/L (0.55 IU/mL), the BioPool TT was >150 s; at 600 IU/L (0.6 IU/mL), the BioMerieux reagent produced a TT of 58 s, and the Sigma Diagnostics reagent produced a TT of 58 s. Intra- and interrun (total) imprecision (CV) for anti-Xa measurements was <10% and 19%, respectively. The precision of the TT assay was evaluated by performing intra- and interrun precision studies. Table 2 in the data supplement summarizes the imprecision data. The Diagnostica Stago, BioPool, and Sigma thrombin reagents were the most precise (CV <2%), followed by BioData thrombin and the BioMerieux reagent. All CVs were <5%. Diagnostica Stago, Sigma Diagnostics, BioPool, and BioData thrombin reagents were stable for 7 days when stored at 2–8 °C. The BioMerieux reagent was stable for 24 h (data not shown). According to the package insert, the latter reagent is stable for 48 h, but at 48 h, control clotting times had exceeded the upper limit of the established reference interval. Comparison studies for each method were performed by evaluating 30 samples from previous runs, encompassing the entire reportable range. Thrombin concentrations differed among reagents, thus producing different reference intervals. In 2 × 2 contingency tables, the percentage of agreement of the “gold standard” method with the new reagents was 93% for each reagent (data not shown). TT assays are based on the original work of Clauss (2). These assays have been used to diagnose disseminated intravascular coagulation and to screen for heparin, hypofibrinogenemia, or dysfibrinogens (1)(3)(4). Modern laboratory use of this assay has focused on the last three. Diagnostica Stago is the only reagent of human origin; all others are of bovine origin. It should be noted that acquired inhibitors may appear in patients with previous exposure to bovine thrombin (5). Such patients will have abnormal TTs when the assay is performed with a bovine thrombin reagent. For a reference laboratory, stability is an important aspect to take into consideration when choosing among reagents. Reconstituting a fresh vial of reagent daily dramatically increases costs when the majority of a previous vial is discarded. All reagents tested were relatively inexpensive, costing an average of US $3.00/mL (list price). In summary, all assays detected hypofibrinogenemia, although the BioMerieux reagent yielded slightly prolonged values for the normal sample on one day. All reagents were sensitive to heparin, although the BioData reagent was too sensitive, with a terminal clotting time at heparin concentrations >100 IU/L. Use of this reagent would make it difficult to monitor a patient receiving therapeutic amounts of heparin. In contrast, the BioMerieux, BioPool, and Sigma Diagnostics reagents produced measurable clotting times in the presence of therapeutic heparin. These latter two reagents have potential use in monitoring heparin therapy.