13th Annual Conference of the British HIV Association 29 March–1 April 2007, Brighton, UK

IS SIMPLIFICATION FROM PI-BASED TO NNRTI-OR NRTI-BASED HAART ASSOCIATED WITH REDUCED IMMUNE-RECOVERY?

HIV Med 2007; 8(Suppl. 1):10 (abstract no. P3)

Giuseppe Lapadula¹, Carlo Torti¹, Sundhiya Mandalia², Anton Pozniak², Antonella d’Arminio Monforte³, Antonella Castagna⁴, Cristina Mussini⁵, Andrea Antinori⁶, Paola Cicconi³ and Giampiero Carosi^1
1Institute for Infectious and Tropical Diseases, Brescia, Italy, ²Chelsea and Westminster Hospital, London, UK, ³S. Paolo Hospital, Milan, Italy, ⁴Vita-Salute University, Milan, Italy, ⁵University of Modena and Reggio Emilia, Modena, Italy, ⁶INMI Spallanzani, Rome, Italy

AIM/OBJECTIVE: To assess CD4 recovery after simplification from PI to NNRTI or NRTI.

METHODS: Patients (Pts) from 5 HIV cohorts with viral load below the limit of detection (<500 c/mL, VLBLD) at month 12 of first line PI-based HAART were grouped as: (G1) Pts switched to 2NRTI+NNRTI from month 12 thereafter (if VLBLD); (G2) Pts switched to 3–4 NRTI (if VLBLD); (G3) Pts continued on PI. Follow-up was right censored in case of loss to follow-up, death, viral failure or treatment discontinuation. Linear mixed models (LMM) were used to derive time weighted CD4 changes over time since switch (G1 and G2) or from month 12 on (G3). The rate of CD4 change occurring in the 12 months pre-switch in G1 and G2 was compared to rate during the 12 months post-switch.

RESULTS: A total of 803 pts were included in G1, 336 in G2 and 891 in G3 with a median follow-up of 29 (IQR: 12–50), 19 (3–37.5) and 17 (8–34) months, respectively. 79% of pts were male and 16% had a boosted PI included in their first HAART. Median CD4 at switch was 439.5 cells/mm³. Using LMM, higher CD4 increase was detected in G3 (estimated mean increase: +2.66 cells/mm³/month; 95% CI 2.43–2.89) than in G1 (+2.13; 1.94–2.32) and G2 (+1.89; 1.53– 2.25). The difference in overall linear trends amongst G3 with either G1 or G2 was statistically significant (P=0.0005 and P=0.0004, respectively). A multivariate LMM showed similar findings after adjusting for residual or confounding effects of cohort, age, gender, nationality, risk factor for HIV, calendar year, use of boosted PI in first line regimen, use of AZT, HCV-Ab serostatus, CDC classification, CD4+ slope prior to study enrolment and time since HAART initiation). The estimated rate of CD4 increase 12 months pre-switch to NNRTI (+7.9 /month; 95%CI 7.2–8.7) was higher than that of 12 months post- switch (+4.4; 3.5–5.3). Similarly, the rate decreased after switching from PI-to NRTI-based regimen [+7.1 (95% CI 5.9–8.4) and +3.9 (2.2–5.6) 12 months before and after the switch, respectively].

CONCLUSIONS: Immune recovery after simplification to NNRTI or NRTI in presence of VLBLD was statistically but modestly lower than that of patients who continued on PI. Switching to these regimens seemed to decrease the immunological benefit derived from HAART, but the clinical significance of this effect is debatable. Further studies are necessary to investigate whether PI-based therapy is associated with improved immunological recovery when compared with NNRTI-based therapy, despite virological equivalence.

2007-03-29
P3

Copyright © 2007 - British HIV Association (BHIVA) Reproduction of this abstract (other than one copy for personal reference) must be cleared through the BHIVA Organising Secretariat 1 Mountview Court, 310 Friern Barnet Lane, London N20 0LD