Incidence and Natural Course of Portal Vein Thrombosis in Cirrhosis

Abstract

To the Editor: We have read with interest the recent article by Maruyama et al. (1) on the incidence and natural course of portal vein thrombosis (PVT) in viral cirrhosis. Despite the interest of the study being the third to report incidence of PVT in cirrhotics, it presents several inconsistencies and drawbacks in design and data analysis. First, reported incidence of PVT included six patients with only splenic vein thrombosis, thus being 36/150, 24%. Moreover, the occurrence of PVT is strangely not equally distributed during the follow-up, with a 12.8% incidence in the first years and about 2% per year during the follow-up. This is in contrast with the current knowledge of a greater prevalence of PVT in patients with more advanced liver disease. Second, patients were recruited for the study during 11 years, and it is stated that Doppler ultrasonography (US) were performed at least every 6–12 months. However, surprisingly, the median number of US per patient was only 4.4, leading to a period of time covered of a maximum of 4 years. Third, correlation of flows at US Doppler at inclusion with occurrence of PVT after years from the initial evaluation is unlikely due to changing hemodynamics in cirrhotic patients overtime. Fourth, it is difficult to understand why patients were tested for prothrombotic disorders at the beginning of the enrollment in the study (in year 1998 the potential role of these alterations in cirrhosis was still not defined). Fifth, US is characterized by diagnostic limitations, especially in extra-hepatic segments of PV or other splanchnic veins. Sixth, worsening of PVT has been described in only 1/23 patients, data very different from previous studies that described progression of PVT in 48–75% of cases at 2 years (2,3). Finally, the way that data on PVT and survival was analyzed does not allow to conclude that PVT does not have an impact on survival. Kaplan–Meyer curve of survival is faulty because the number of patients plotted in the figure is 156 and not 150, and survival was considered from the enrollment in the study. Therefore, for example a patient developing PVT after 7 years of follow-up and dying 1 month after it is considered as being alive for 7 years and 1 month. In conclusion, we believe that the interpretation of the described data is difficult and no recommendations can be drawn according to the results of the present manuscript.