Background:  Treating HIV-1-infected infants soon after birth remains a crucial issue. Guidelines from different countries differ in whether to start immediate or differed therapy in asymptomatic children <12 months of age. In PENTA 7, 20 infants received didanosine (ddI) + stavudine (d4T) + nelfinavir (NFV) before 3 months of age. Clinical and immunological benefit was seen at 18 months. However virologic failure and emergence of resistance was a major concern. We describe long-term response to 5 years.

Methods:  We followed 19 children (1 died at week 60 of non-HIV causes) beyond 18 months. HIV events, CD4 cell counts and percentages, HIV-1 RNA, ART received, and weight and height measurements were collected every 3 months between 18 months and 3 years and annually beyond 3 years. Changes in CD4, height and weight z-scores from baseline were assessed using the paired Wilcoxon signed rank test.

Results:  Median follow-up was 5.2 years (IQR 5.0 to 5.9). All children, but 1, were seen at least at 5 years (consent was withdrawn after 3.8 years of follow-up). No AIDS events were reported; 2 children were still on initial therapy and 4 had stopped ART permanently; and 13 children switched to second- (11) or third-line (2) regimen using 2 new drugs (6 children), 3 (5), or 4 (2) including zidovudine (ZDV) (2), lamivudine (3TC) (10), abacavir (ABC) (5), emtricitabine (FTC) (1), lopinavir/ritonavir (LPV/r) (9), efavirenz (EFV) (4), or nevirapine (NVP) (4). Changes were due to lack of viral load response (6), resistance mutations (3), or other (4). Among the 18 children analyzed, 10 had a viral load <400 copies/mL (7 <50 copies/mL). Median viral load of the 8 remaining children was relatively low at 2570 copies/mL (range 479 to 44,300 copies/mL). The proportion of children <400 copies/mL (56%) was similar to that observed at 18 months (47%). Median CD4% was 31% (range 22 to 44), with median increase from baseline in CD4 z-score +0.66 (IQR –0.9 to 1.5, p = 0.19). Weight and height adjusted for sex and age had increased significantly by 5 years (median increase z-score from baseline +0.90 and +0.98, respectively, both p = 0.002).

Conclusions:  By 5 years, children did not show any major clinical progression, maintained good level of CD4 percentage, and stable viral load. Growth improvement found at 18 months was sustained through long-term follow-up. While most of the children switched to second- or third-line regimen, new resistance mutations were reported in few children. Starting HAART early in life with more potent, suitable, and palatable drugs may prevent early switches and thus preserve future drug options.