Splenomegaly--an insensitive sign of portal hypertension

The prevalence of splenomegaly associated with portal hypertension was examined in a consecutive population of 111 patients who had portal hypertension diagnosed using specific endoscopic, sonographic, and Doppler signs. Splenic size was measured objectively via its cranio-caudal length on coronal section using ultrasound and by clinical examination. Sonographically, 52% of patients had a definitely large spleen and 35% a spleen less than one standard deviation from the normal mean, while a further 13% had equivocal splenomegaly. Only 52% of patients had splenomegaly on clinical assessment. Splenomegaly was less common in patients with alcoholic (41% definite, 15% equivocal) than in those with non-alcoholic liver disease (66% definite, 17% equivocal, p = 0.02) and splenic length was significantly smaller in alcoholic patients (12.7 +/- 0.5 cm) compared to patients with either non-alcoholic liver disease (15.0 +/- 0.6 cm, p = 0.003) or portal hypertension due to vascular occlusive diseases (16.5 +/- 2.0 cm, p = 0.006). Splenomegaly, whether assessed sonographically or clinically, is an insensitive sign of portal hypertension and its absence cannot be used as a negative predictor of the presence of portal hypertension in patients with chronic liver disease.