Idiopathic non-cirrhotic portal hypertension (NCPH) is caused by obliterative venopathy affecting the terminal radicals of the hepatic portal vein. It has been associated with areas of low socio-economic development. We observed a long history of thrombocytopenia and association with enteric disorders in a Caucasian population presenting with NCPH and postulated that the underlying mechanism could be deficiency of ADAMTS13, leading to the circulation of ultra large molecular weight von Willebrand factor (VWF) multimers and consequent platelet aggregate formation, causing occlusion of hepatic portal venules. ADAMTS13 activity was measured in all patients by a collagen binding assay, while antigen and IgG antibody were measured using ELISA kits (American Diagnostica Inc), normal ranges: 66–126%, 485–1242ng/ml and <11AU/ml, respectively. Plasma ADAMTS13 activity and antigen levels were measured in 14 patients with NCPH and 25 control patients with miscellaneous liver diseases associated with portal hypertension and Model for End Stage Liver Disease (MELD) scores indicating similar disease severity, median 12 (IQR 10–14) and 11 (8–16), for NCPH cases and controls. The platelet count was reduced in 13/14 NCPH cases median 69 x109/L (IQR 40–97) and 123 (85–193) in cases and controls, but did not correlate with MELD score or ADAMTS13 activity. ADAMTS13 activity was very low in 13/14 NCPH cases (median 12% (IQR <5–20)) and significantly lower than controls (59% (4–84), p<0.0001). In 4 cases, serial samples over several months revealed consistently low ADAMTS13 levels. There was a corresponding reduction in ADAMTS13 antigen, median 238ng/ml (IQR 186–361) and 584 (442–831), although NCPH cases showed a small excess of antigen over activity (antigen:activity ratio: median 25.7 (IQR 11.3–32.6) and 9.4 (8.1–13.8), in cases and controls, p<0.001), suggesting that an abnormal molecular species could be present. An inhibitor was not detected, although 10 cases and 15 controls had IgG anti-ADAMTS13: median levels 14.8AU/ml (8.8–20.0) and 13.4 (9.0–22.6), but these IgG levels did not correlate with activity. Non-inhibitory IgG antibodies to ADAMTS13 have been previously observed in certain other clinical groups (eg. coronary syndromes, sepsis, etc.). In NCPH, there was no correlation between ADAMTS13 level and either VWF antigen or VWF collagen binding activity. Increased very high molecular weight VWF multimers were observed in 4/11 NCPH cases and 0/22 controls. CD31 positive platelet aggregates were observed in the hepatic vessels of 7/20 (35%) NCPH cases and 1/15 (7%) controls. Our results suggest that sustained deficiency of ADAMTS13 associated with thrombocytopenia is characteristic of NCPH. It differs from other hepatic pathologies, where the decrease in ADAMTS13 has been reported to correlate with the severity of liver disease. NCPH also differs from thrombotic thrombocytopenic purpura because there is a chronic disease process, there is an absence of obvious red cell fragments and the primary organ affected is the liver. Since the predominant plasma source of ADAMTS13 appears to be hepatic stellate cells, the portal microcirculation is most at risk from pathological interactions of VWF and platelets if enteric factors further deplete circulating levels of ADAMTS13. The cause of the decreased ADAMTS13 remains to be elucidated, but could be due to decreased expression and synthesis, immune depletion (IgM and IgA anti-ADAMTS13 have not been investigated), or consumption in the circulation.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author