I read with interest the editorial by Dr. Marlar1 that accompanied the two articles2, 3 on the standardization of the prothrombin time used to calculate the Model for End-Stage Liver Disease (MELD) score to prioritize patients for liver transplantation.4 The fact that two different articles reached the same conclusion strongly indicates that the regular international normalized ratio (INR) is not suited to harmonizing results for patients with chronic liver disease.5 The current INR is not valid for this purpose because it is calculated with the international sensitivity index (ISI), which is currently determined and validated only for patients on vitamin K antagonists (VKA).6 The use of the same ISI value to calculate the INR for the two settings would be possible only if the coagulation defect induced by chronic liver disease were the same as that induced by vitamin K antagonist treatment. Numerous theoretical considerations point to the conclusion that liver failure and anticoagulation with vitamin K antagonists do not have the same effect on the coagulation process.7 Therefore, two different ISI values are needed to harmonize results in the two settings: the regular ISI (that is, ISIvka), calculated as recommended by the World Health Organization with plasmas from patients stabilized on vitamin K antagonists,6 and the ISIliver, calculated as suggested by the authors of the two articles2, 3 with plasmas from patients with chronic liver disease.
The point at issue is how to provide the clinical laboratory with the two values. Presently, the ISIvka is determined and provided by manufacturers of commercial thromboplastins that comply with the recommendation issued by the World Health Organization6 that the “manufacturer of thromboplastins should state on the package insert the ISI of the relevant batch of thromboplastin together with the reference preparation against which it has been determined and the instrument for which it is valid.” Thus, the ISIliver could also be determined and provided by the manufacturer. Although possible, it will presumably take time before this can be fully implemented. Bellest et al.3 suggested a local determination of the ISIliver by means of frozen plasmas from patients with liver failure to which the certified INR for cirrhosis could be assigned by an external authority. This system of local calibration could be derived from a modification of a similar system recommended by the working group of the Subcommittee on Control of Anticoagulation of the International Society on Thrombosis and Haemostasis.8 However, even this system may be difficult to implement. First, the plasmas should be prepared in large amounts to be distributed to individual clinical laboratories. Second, frozen plasmas, although suitable for ISI calibration, are probably not stable for years as required for any reference material. Lyophilization would certainly increase shelf life but may add undesirable between-thromboplastin variability because of the alteration of coagulation proteins consequent to freeze drying, the addition of preservatives, or both.8 Third, it would be difficult to identify an independent authority that should be responsible for the preparation and certification of those reference plasmas.
Alternative solutions could be the use of recombinant human or combined ox brain thromboplastins for INR measurement. These thromboplastins could probably be used to calculate the INR in the setting of chronic liver disease with either the ISIvka or the ISIliver. The two recombinant thromboplastins (Recombiplastin and Innovin) and the combined ox brain thromboplastin (Thrombotest) included in our study2 showed a negligible difference between the ISIvka and the ISIliver that amounted to 8.6%, 11.8%, and 5.7%, respectively.2 Those ISI differences translated into small INR or MELD differences.2 Interestingly, the correspondent ISI difference observed for one of the human recombinant thromboplastins (that is, Recombiplastin) included in the study of Bellest et al.3 amounted to 2.5%.3 It seems, therefore, that the aforementioned thromboplastins display similar ISIvka and ISIliver. Although promising, these observations need to be confirmed in other studies and with different coagulometers.
A more pragmatic approach could be to identify a few laboratories within the United States, for example, that could assist the United Network for Organ Sharing as reference laboratories for INR measurement in patients with chronic liver disease listed for liver transplantation. Those laboratories could be conveniently identified among those which use the same thromboplastin/coagulometer combinations, for which the ISIliver could be easily determined according to the protocol described in the two articles.2, 3 The laboratories could also act as references for the measurement of the other two parameters needed for the MELD calculation (that is, bilirubin and creatinine), which are (probably) not devoid of standardization problems. This pragmatic approach would have the advantage of minimizing the interlaboratory variability of the MELD, thus making the score a true “objective” parameter for allocating organs.4 The inherent disadvantage would be the need for centralization of plasma samples.