Background: Viral load is the most important marker of response to antiretroviral therapy (ART). Decreases in viral load following ART initiation are associated with decreases in mortality and HIV disease progression. Early ART initiation provides added mortality advantages, but it has not been clearly demonstrated whether this is related to differences in rates of virologic failure.
Methods: 816 HIV-infected treatment-naïve Haitian subjects with CD4 counts between 200-350 cells/mm3 were randomized to early (within 2 weeks; N=408) versus delayed ART initiation (CD4 count ≤200 cells/mm3 or development of an AIDS-defining condition; N=408) between 8/1/2005 and 7/31/2008. In the early arm, 340 subjects had viral load data available at 2 years. In the delayed arm, 23 subjects died and 30 were lost prior to ART initiation. Of the 355 subjects in the delayed arm that started therapy, 274 had viral load data available at 2 years. A validated 9-item food security questionnaire was administered at ART initiation, and all subjects were then scheduled for follow-up with a study physician at 1, 2, 3, 6, 12, 18 and 24 months; during these visits ART adherence was assessed using 3-day recall. All analyses were performed using stata 13 software (StataCorps, TX).
Results: At ART initiation, subjects in the delayed arm (N=355) had more advanced HIV disease (33.7% vs. 19.10% WHO Clinical Stage III/IV; p-value <0.001) and greater food insecurity (74.1% vs. 54.8% scoring 5-9; p-value <0.001). After 2 years on therapy there was no difference in virologic failure (viral load >50 copies/μL) between the arms (21.6% early vs. 22.3% delayed). Table 1 shows that in multivariate analysis the predictors of virologic failure were younger age (OR per additional year 0.98; 95%CI: 0.96-0.99, p-value 0.014), annual income less than US$100 (OR 1.56, 95%CI: 1.06-2.31, p-value 0.026), suboptimal adherence (OR 1.81, 95%CI: 1.25-2.63; p-value 0.002) and number of visits attended (OR 0.68, 95%CI: 0.57-0.82, p-value <0.001).
Conclusions: Delaying ART initiation until CD4 count ≤200 cells/mm3 or development of an AIDS defining condition resulted in significant pre-ART attrition, interim progression of HIV disease and greater food insecurity. Delayed ART initiation was not associated with increased virologic failure. Predictors of virologic failure were related to age, income and adherence to therapy.
| Univariate | Multivariate Model 1 | Multivariate Model 2 | |
| Odds Ratio (95%CI, p-value) | Odds Ratio (95%CI, p-value) | Odds Ratio (95%CI, p-value) | |
| Early ART initiation arm | 0.89 (0.62-1.27; 0.515) | 0.97 (0.66-1.41; 0.867) | Excluded |
| Age at ART initiation | 0.97 (0.96-0.99; 0.008) | 0.98 (0.96-0.99; 0.031) | 0.98 (0.96-0.99; 0.014) |
| Female | 1.14 (0.79-1.63; 0.48) | Excluded | Excluded |
| Marital status | |||
| * Married | 1 (reference) | 1 (reference) | Excluded |
| * Unmarried/ widowed | 1.79 (1.09-2.97; 0.022) | 1.60 (0.95-2.70; 0.077) | Excluded |
| Food Security Score | |||
| * 0 – 3 | 1 (reference) | Excluded | Excluded |
| * 4 – 6 | 1.46 (0.73-2.91; 0.283) | Excluded | Excluded |
| * 5 – 9 | 1.49 (0.97-2.29; 0.072) | Excluded | Excluded |
| Annual income | 1.57 (1.07-2.29; 0.020) | 1.55 (1.05-2.30; 0.029) | 1.56 (1.06-2.31; 0.026) |
| Literate | 0.88 (0.60-1.29; 0.508) | Excluded | Excluded |
| WHO Clinical stage at ART initiation | |||
| * Stage I | 1 (reference) | Excluded | Excluded |
| * Stage II | 1.08 (0.70-1.65; 0.739) | Excluded | Excluded |
| * Stage III or IV | 1.07 (0.65-1.76; 0.782) | Excluded | Excluded |
| Tuberculosis at ART initiation | 2.25 (1.12-4.55; 0.023) | 1.84 (0.87-3.90; 0.111) | Excluded |
| Suboptimal adherence | 1.91 (1.33-2.74; <0.001) | 1.79 (1.23-2.61; 0.002) | 1.81 (1.25-2.63, 0.002) |
| Number of visits | 0.77 (0.66-0.89; 0.001) | 0.69 (0.57-0.84; <0.001) | 0.68 (0.57-0.82; <0.001) |