We read with very great interest the prospective study1 from the European Network for Vascular Disorders of the Liver (EN-Vie) providing stronger evidence of anticoagulation for acute portal vein thrombosis (PVT) without cirrhosis than any previous retrospective studies with a smaller sample size.2–4 Despite the other clear and valuable results in the study, existence of a statistically significant difference between recanalization rate of portal vein and that of splenic or superior mesenteric vein is confusing (Table 1), and is marginal in appearance, which seems to be attributed as a plateau in 1-year recanalization over time for portal vein (recanalization did not occur beyond 6th month), but there was no apparent plateau for splenic and superior mesenteric vein (recanalization stopped in the 9th and 11th month, respectively). Is the discrepancy rational between different segments of the same portal vein system? If so, why?
|Location of Obstruction||Present* (%)||Absent† (%)||P Value‡|
|Portal vein or both its two branches||58 (70%)||29 (30%)||0.032|
|Superior mesenteric vein||19 (47%)||25 (53%)|
|Splenic vein||22 (46%)||26 (54%)|
Perhaps one of the major reasons lies in the degree of thrombosis—complete or partial—that is missing among factors predicting recanalization, which unanimously exists in the four classical studies1–4 about anticoagulation for PVT. Naturally, the difficulty of recanalization increases with degree of thrombosis, whether acute or chronic in stage and noninvasive or invasive in management of PVT. In other words, the gap among patency of various portal venous segments would have disappeared, if the data were stratified according to complete or partial obstruction.
In contrast, degree stratification is consistently involved in the studies about outcome of invasive therapies for PVT5–9 due to its close associations with operability and prognosis. Herein, diverse classifications of PVT were listed in Table 2, except for simple classification into complete and partial thrombus. Further, some classical images in our patients are demonstrated (Fig. 1) for a clear distinction between partial and complete occlusion, which are mainly characterized by partial and complete absence of flow within portal vein on color Doppler ultrasound or angiography, or a filling defect and “train track” appearance of enhancement on computed tomography.10
|Authors (published date)||Grade I||Grade II||Grade III||Grade IV|
|Stieber et al. (1990)||Partial PVT or patent PV||Complete PVT ± SVT||Extensive thrombosis (including SMV)||NA|
|Nonami et al. (1992)||Thrombosis within intrahepatic portal branches||Thrombosis within RPB or LPB or the bifurcation||Partial thrombosis within MPV||Complete thrombosis within MPV|
|Yerdel et al. (2000)||<50% PVT ± minimal obstruction within SMV||>50% PVT ± minimal obstruction within SMV||Complete PV and proximal SMV thrombosis||Complete PV and entire SMV thrombosis|
|Jamieson NV. (2000)||Thrombosis confined to PV beyond the splenomesenteric confluence, which may be either partial or complete||Thrombosis extending into the proximal SMV system, but with a patent vessel in the mesentery||Diffuse thrombosis of the splanchnic venous system, but with large accessible collaterals||Extensive thrombosis of the splanchnic venous system, but with only fine collaterals|
|Bauer et al. (2006)||Less than 25% occluded in PV, SMV or SV||26%–50% occluded in PV, SMV, or SV||51%–75% occluded in PV, SMV or SV||76%–100% occluded in PV, SMV, or SV|
From our perspectives, it is necessary for prediction of recanalization to accurately distinguish between complete and partial thrombosis in any study on management of PVT.