Risk factors and outcome of HIV-associated idiopathic noncirrhotic portal hypertension

Background Idiopathic noncirrhotic portal hypertension (INCPH) has been reported increasingly in patients with HIV infection. Aim To evaluate the number of nationwide diagnosed HIV-associated INCPH cases and to assess its clinical features, risk factors and outcome. Methods All HIV centres in the Netherlands were contacted and requested to notify INCPH cases diagnosed in their population. A case-control study was performed to identify the risk factors of INCPH. The cases were group-matched for duration of follow-up after HIV diagnosis to controls. Controls were selected from a database of HIV patients with negative screening for signs of portal hypertension on abdominal ultrasound. Univariate and multivariate conditional logistic regression analyses were performed. Results On 1st of July 2011, 18.085 individuals were infected with HIV in the Netherlands. Within this population, sixteen patients with clinically overt INCPH were identified. At the time of INCPH diagnosis, cases had a lower platelet count and a higher ALT level. In univariate and multivariate analyses, didanosine [OR: 1.9 (1.3-2.8)], concomitant didanosine and stavudine treatment [OR: 6.3 (2.1-19.1)] and concomitant didanosine and tenofovir treatment [OR: 5.1 (1.2-22.6)] were independently associated INCPH. During follow-up, 4 patients died [malignancy (n = 3), liver failure (n = 1)]. A significant decline in platelets was observed after didanosine discontinuation (P = 0.003). Conclusions HIV-associated clinically relevant idiopathic noncirrhotic portal hypertension appears to be a rarely diagnosed disease. Long-term exposure to didanosine and short-term combination of didanosine and stavudine or tenofovir exposure are associated with idiopathic noncirrhotic portal hypertension. Mortality in HIV-associated idiopathic noncirrhotic portal hypertension is mainly related to HIV-associated disorders. Portal hypertension continues despite didanosine discontinuation