Tempo di protrombina: quale denominatore per il PT Ratio e quali possibilità di utilizzo dell'INR al di fuori della terapia anticoagulante orale

Background: In routine laboratory practice, the prothrombin time (PT) is used either to detect acquired or congenital deficiencies of the clotting system or to monitor anticoagulant therapy with vitamin K antagonists. Expression of results as INR is recommended only for the latter case; still in many laboratories INR values are reported irrespectively of whether plasma samples are from patients on oral anticoagulation. In addition, for the denominator term in the expression of PT results, some laboratories use commercial lyophilized normal plasma instead of the mean normal prothrombin time (MNPT), which may result in different normal ranges with the same reagent-coagulometer combination. Methods: On the same coagulometer (STA-R, Roche), the imprecision of two thromboplastin reagents (Neoplastin Plus, rabbit brain, Roche, coagulometer-specific ISI 1.31, and Innovin®, recombinant human tissue factor, Siemens) was tested with reconstituted lyophilized normal and abnormal plasmas (Roche, control plasma 1 and 2; Siemens, normal and abnormal plasma and INR calibrator L1). The normal range of PT ratios with the two thromboplastin reagents was tested in 40 apparently healthy subjects using either the MNPT, previously determined in a different series of 20 volunteers, or the lyophilized normal plasmas as denominator term. A comparison of the two modalities of PT ratio expression with the two reagents was conducted on 200 plasma samples from out- (n=90) and in-patients (n=110) not on oral anticoagulation. Results in patients were also compared with the INR modality of expression. Results: Total imprecision (CV%) of the two reagents, tested with reconstituted normal and abnormal plasmas, was similar and ranged from 1.1% to 2.5%. In controls, average PT ratios were similar with the two thromboplastin reagents when using the MNPT as denominator term and they did not differ significantly from 1.0 (p ≥0.17). With the normal lyophilized plasmas as denominator term, they consistently differed from 1.0 (from 0.90±0.07 to 1.05±0.08, p ≥0.03). In patients, average PT ratios were similar only when using the MNPT as denominator term. According to a simplified WHO procedure and assuming Neoplastin Plus as the reference thromboplastin, the coagulometer-specific SI of Innovin® was 0.77 (95% confidence limits: 0.73-0.80), with Tomenson's correction (10 ^-d) equal to 1.09. When PT ratios of patients were transformed into INR values, these differed significantly irrespective of the denominator term, and the percentage of patients with abnormal prothrombin time values with Innovin® increased from 38.5 to 62.5. Conclusions: Commercial "normal" plasmas are no substitute for the MNPT in determining normal ranges for the prothrombin time, and should be used only for the purpose of quality control. Expression of PT results as INR in patients not on oral anticoagulation is misleading and should be avoided in routine laboratory practice.