Methods: A case–control study involving 100 patients with cirrhosis and PHT with (n = 50) or without (n = 50) PSCVs detected by using MSCT. Direct transjugular HVPG was measured in 10 patients with (n = 5) or without (n = 5) PSCVs using Swan-Ganz catheters.

Results: A total of 47(94%) of the patients with PSVCs had SCVs, which were classified according to their site (hilum, upper pole, lower pole), and shape (grape-like, serpiginous, worm-like). Compared to patients without PSVCs, those with PSVCs has significantly less endoscopically-detectable gastroesophageal varices (58% vs. 92%; P < 0.001), more hemorrhoids (48% vs. 20%; P < 0.001), higher rate of HCC (88.2% vs. 11.8%; P < 0.001), more HE recurrence (4.9 ± 2.4 vs. 2.8 ± 1.62% attacks in 6 months; P < 0.001), lower HVPG (2.2 ± 1.6 vs. 12.8 ± 1.6 mmHg; P < 0.001), but similar splenic size (14.7 ± 2.8 vs.14.8 ± 0.9 cm; P = 0.788), respectively.

Conclusions: MSCT accurately delineates PSVCs, and SCVs are the most common. Clinical significance of SCVs classification needs further investigated. Patients with PSVCs may require less variceal endoscopic surveillance, more HCC screening, and prolonged HE prophylaxis.