Results
Population
We included 3,338 patients from 14 different cohorts in this analysis (Figure 1). The number of patients per cohort ranged from 46 to 891, with nine cohorts from sub-Saharan Africa (accounting for 80.1% of the overall study population), three cohorts from Latin America (14.7%), one cohort from Morocco and one from India. The patient group had the following characteristics: 2,139 patients were female (64.1%), the median age was 34 years (inter-quartile range (IQR) = 30–41); 32.3% of the patients were in WHO stage I or II at ART initiation; 2,070 (62.0%) patients initiated their first-line ART in or after 2004; and for 96.0% of the patients, the ART regimen included an NNRTI and two NRTIs as per WHO recommendations. The follow-up characteristics of the patient group are as follows: the total number of person–years of follow-up was 6,778; the median follow-up time was 1.6 years (IQR = 1.2–2.4); 3,039 patients (91.0%) were still followed up 12 months after ART initiation and 1,995 (59.8%) were still followed up at 18 months; and 44 (1.3%) and 325 (9.7%) patients were lost to follow-up at 12 and 18 months, respectively (Table 1).
(Enlarge Image)
Figure 1.
Study flow chart. ART-LINC of IeDEA Cohort Collaboration.
CD4 Cell Count and HIV-RNA: Data Description
A majority of patients (83.9%) had three or more HIV-RNA determinations during the study period. Similarly, 87.8% had at least three CD4 cell counts and 46.2% had a greater number of measurements up to 18 months. Median HIV-RNA level (log10) at ART initiation was 5.07 (IQR 4.53–5.54) and the median CD4 cell count was 107 cells/mm (IQR = 46–179); 3,038, 2,969 and 814 individuals had an HIV-RNA level below the threshold of 500 copies/ml at M6, M12 and M18, respectively, after starting ART (91.0%, 89.0% and 87.7% of all available viral load measurements, respectively).
Among patients initiating ART with HIV-RNA levels above 10,000 copies/ml, 91.7% (n = 1,780) and 90.0% (n = 1,747) achieved HIV-RNA levels below 500 copies/ml six months and 12 months, respectively, after ART commencement. Among patients with undetectable (<500 copies/ml) HIV-RNA levels (n = 3,038) at M6, 93.0% (n = 2,825) remained with an undetectable HIV-RNA level and 3.0% (n = 92) had an HIV-RNA level ≥10,000 copies/ml at M12. Among patients with undetectable HIV-RNA levels six and 12 months after starting ART (n = 2,825), 750 (94.1% of patients with available data) remained with an undetectable HIV-RNA at M18.
CD4 Cell Count and HIV-RNA Longitudinal Analysis
We assessed the effect of the HIV-RNA level while on ART on the mean CD4 cell count slope. Figure 2-a shows the adjusted mean CD4 cell count slopes (for the 6 to 12 months period) of patients with an M6 HIV-RNA level below or above a threshold of 10,000; Figure 2-c gives these slopes using an HIV RNA threshold of 5,000. Figure 2-b shows the CD4 cell count slopes (for the 12 to 18 months period) of patients with an M12 HIV-RNA level below or above a threshold of 10,000; Figure 2-d shows the same slope for patients with a 5,000 threshold. The CD4 cell count SLOPE was not significantly different from the null value between six and 12 months following ART initiation, regardless of the HIV-RNA replication level (Figures 2-a, c). However, differences related to HIV-RNA replication levels were observed after 18 months: only patients with suppressed HIV-RNA levels at 12 months had a continuous increase in CD4 cell count, whereas those with any degree of HIV-RNA replication (either 10,000 or 5,000 copies) 12 months after ART commencement had either no increase or a decrease of their CD4 cell count slope (Figures 2-b, d).
(Enlarge Image)
Figure 2.
a: Mean CD4 cell count change between M6-M12 relative to the M6 viral load (adjusted, threshold = 10,000 copies/ml). b: Mean CD4 cell count change between M12-M18 relative to the M12 viral load (adjusted, threshold = 10,000 copies/ml). c: Mean CD4 cell count change between M6-M12 relative to the M6 viral load (adjusted, threshold = 5,000 copies/ml). d: Mean CD4 cell count change between M12-M18 relative to the M12 viral load (adjusted, threshold = 5,000 copies/ml).
Table 2 shows that age, sex, CDC stage, and the calendar year in which ART was initiated were all significant determinants of CD4 cell count six months after ART initiation. Moreover, we modelled the effect of several HIV RNA values at M6 and M12 following ART initiation on the CD4 cell count changes between M6 and M18; CD4 cell count increased by 90 cells (95% CI = 71; 108) per year when patients had a suppressed HIV-RNA level at both M6 and M12; in contrast, CD4 cell count decreased by 1 cell per year (CI = −51; 48) when HIV-RNA was consistently above 10,000 copies (Table 2). In addition, similar CD4 cell count changes were observed when a lower HIV-RNA threshold (5,000 copies) was chosen as a cut-off (AIC = 106669 for the 10,000 versus 106679 for the 5,000 copies threshold) [data not shown].