Is CD4 percentage a better marker of immune-suppression than CD4 absolute count in HIV-infected subjects with cirrhosis?
Objectives: The relative prognostic value of CD4 percentage and absolute count to predict AIDS-events has not been evaluated in HIV-infected, cirrhotic subjects.
Methods: We evaluated the incidence of AIDS-events in subjects with and without cirrhosis or liver disease in an Italian cohort of HIV-positive, antiretroviral-naive subjects (I.Co.N.A.). Three groups were identified: Group-1, subjects with cirrhosis (histological and/or clinical); Group-2, HCV and/or HBV co-infected subjects without cirrhosis; Group-3, subjects without liver diseases (HCV and HBV-negative, ALT<40 IU/L). The incidences of new AIDS-events and the relative prognostic values of CD4 percentage and absolute count were evaluated by Poisson regression model.
Results: 6,126 subjects were studied (30% females, median age 35 years); CD4 count and percentage at enrolment were 427 (1-1361) and 23% (1%-81%); HIV-RNA was 4.3 log10 copies/mL (1.3-6.8). The prevalence of HCV-Ab and HBsAg positive subjects was 38% and 5%. 402 new AIDS-events were reported during 25,693 PYFU from the date of enrolment to AIDS diagnosis or to last available clinical follow-up (incidence: 1.6x100 PYFU, 95% CI:1.4-1.7). In cirrhotic subjects, higher CD4 cells but not higher percentages were associated with a lower risk of AIDS (RR:0.58, 95% CI:0.39-0.85, p=0.006 per 100 cells/mm3 higher; RR:0.61, 95% CI:0.30-1.24, p=0.17 per 10% higher). In Group-2 and Group-3 both absolute CD4 number and percentage were associated with a higher risk of AIDS (Group-2, RR:0.75, 95% CI:0.67-0.83, p=0.0001 per 100 cells/mm3 higher; RR:0.54, 95% CI:0.43-0.67, p=0.0001 per 10% higher; Group-3, RR:0.52, 95% CI:0.42-0.63, p=0.0001 per 100 cells/mm3 higher; RR:0.61, 95% CI:0.43-0.87, p=0.006 per 10% higher).
Conclusions: Our data indicate that the absolute number of CD4 cells is a better predictor of AIDS-events than CD4 percentage in HIV-infected patients with cirrhosis but not in those with HBV and/or HCV co-infection or without liver disease. Absolute CD4 counts remain the better parameter to be consider to start HAART in this population.
Bongiovanni M.1, Gori A.1, Cozzi Lepri A.2, Antinori A.3, De Luca A.4, Pagano G.5, Chiodera A.6, Puoti M.7, D'Arminio Monforte A.1
Clinic of Infectious Diseases, San Paolo Hospital, University of Milan, Italy