Analysis of prognostic predictors and differences between subgroups of pulmonary arterial hypertension associated with portal hypertension and HIV infection

Abstract

Abstract Background Pulmonary arterial hypertension (PAH) represents a complication of both portal hypertension (Po-PAH) and HIV infection (HIV-PAH). Due to overlapping risk factors, it is frequent to have an overlap PAH form associated with both portal hypertension and HIV (HIV/Po-PAH). Methods We enrolled 204 patients with Po-PAH (128), HIV-PAH (41) and HIV/Po-PAH (35), with the aim of comparing the characteristics and identifying the prognostic determinants. At baseline, in patients with Po-PAH and HIV/Po-PAH, the severity of liver disease was estimated using the Child-Turcotte-Pugh (CTP) score, MELD score (Model for End-stage Liver Disease); instead, in patients with HIV-PAH and HIV/Po-PAH it was defined whether they received HAART (Highly Active Anti-Retroviral Therapy) and the CD4+ lymphocytes number. Data are expressed as median (interquartile range) and compared with Dunn’s test. Prognostic determinants were identified by univariate and multivariate Cox-regression analysis. Survival was assessed using Kaplan Meier method. Results HIV-infected patients [40 (37-43) years] and HIV/Po-PAH [44 (39-48) years] are younger than Po-PAH patients [53 (44-60) years]; the exercise capacity assessed by the 6-minute walking test of patients with HIV/Po-PAH [516 (432-571) m] was superior to both patients with Po-PAH [426 (366-504) m] and HIV-PAH [448 (370-510) m], presumably justified by the demographic and haemodynamic differences between these groups. In fact, patients with concomitant portal hypertension (Po-PAH and HIV/Po-PAH) have a better haemodynamic profile than patients with HIV-PAH, specifically, lower pulmonary vascular resistance [6.2 (4.9-8.8) and 7.0 (5.5-11.2) WU vs 8.9 (7.5-12.2) WU] and higher cardiac index [3.1 (2.6-3.8) and 3.0 (2.7-3.7) l/min/m2 vs 2.6 (2.2-2.9) l/min/m2] and mixed venous oxygen saturation [70 (65-75) and 67 (62-71) % vs 61 (55-66)%] values. Despite these differences, the survival of the three groups of patients was not statistically different (5-year survival of patients with Po-PAH was 41%; of patients with HIV/Po-PAH was 46%; and of patients with HIV-PAH was 51%; log-rank test p-value= 0.059) and the prognostic factors that emerge at the multivariate analysis are only parameters associated with the underlying disease. In particular, the parameters independently associated with prognosis are, in patients with Po-PAH, only the CTP score [HR 1.665 (CI 1.153-2.403), p= 0.007]; in patients with HIV-PAH only the treatment with HAART [HR 0.283 (CI 0.119-0.675), p= 0.004]; in patients with HIV/Po-PAH both MELD-Na+ [HR 1.224 (CI 1.064-1.408), p= 0.005] and trans-hepatic venous gradient [HR 1.125 (CI 1.026-1.235), p= 0.013]. Conclusions In patients with Po-PAH, HIV-PAH, and HIV/Po-PAH adequately treated with PAH medications, the prognosis seems to be related to the underlying disease. This may explain why their survival is not statistically different despite significant demographic, exercise and hemodynamic differences.