CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

scores occurred earlier and favored the delayed arm. Based on the results of this study, we recommend that ART initiation in children with SAM should be delayed for at least two weeks after starting nutritional rehabilitation. 818 TREATMENT OUTCOMES OF KAPOSI SARCOMA IN HIV-1–INFECTED CHILDREN IN MOZAMBIQUE Faizana Amodo 1 , Yolanda Cachomba 2 , Dalila Rego 1 , Orvalho Augusto 2 , Dulce Bila 3 , Paula Vaz 3 1 Hosp Central de Maputo, Maputo, Mozambique, 2 Univ Eduardo Mondlane, Maputo, Mozambique, 3 Fundação Ariel Glaser Contra o Sida Pediatrico, Maputo, Mozambique Background: Kaposi Sarcoma (KS) in HIV-1-infected children is not uncommon in sub-Saharan Africa and the outcomes are influenced by the disease stage and respective treatment. Antiretroviral treatment (ART) is now widely accessible but chemotherapy is restricted to a few tertiary centers. Here we present a retrospective of 12 years ‘outcomes of KS in children attending a reference public hospital in Maputo, Mozambique. Methods: Children aged 0-14 years presenting with KS and HIV-1 infection were consecutively recruited, between December 2003 and December 2015, at Maputo Central Hospital. They received ART following WHO recommendations and chemotherapy either Paclitaxel or a combination of Adriamycin, Bleomycin, Vincristine (ABV). Descriptive statistics and survival analysis with competing risks (death, full remission and partial remission) are presented. Follow up ends at event occurrence or censor at 24 months. Results: About 64 children started chemotherapy, with median age 9 years (IQR, 6 - 12) and male to female ratio of 1.7:1. Overall, the median follow-up time was 17 months. At 18th month of follow up, 28 (44%) were alive, 15 (23%) in full remission, 12 (19%) dead, 5 (8%) in partial remission and 4 (6%) lost-to-follow-up. All children were on ART, mostly on a regimen based on 2NRTI+1NNRTi 61 (95%). The incidence of full remission was 8 times higher on ABV chemotherapy when compared to Paclitaxel (p=0.002) (Figure 1). Grade III and IV toxicity was only observed in Paclitaxel users. Conclusion: The present results suggest the superior benefit of ABV use in co-infected KS-HIV-1 children and the importance of early diagnosis of KS. In fact, our results show that despite the use of ABV, if children with KS are diagnosed late the poor the outcome observed.

Poster and Themed Discussion Abstracts

819 DURABLE GAINS IN BONE MASS AMONG SOUTH AFRICAN CHILDREN SWITCHING TO EFAVIRENZ Stephen M. Arpadi 1 , Stephanie Shiau 1 , Renate Strehlau 2 , Faeezah Patel 2 , Ndileka Mbete 2 , Louise Kuhn 1 , Ashraf Coovadia 2 , Michael T. Yin 1 1 Columbia Univ, New York, NY, USA, 2 Empilweni Service and Rsr Unit, Johannesburg, South Africa

Background: We previously reported in our cohort study of pre-pubertal perinatally HIV-infected children that after 2 years, bone mineral content (BMC) Z-score of the whole body (WB) and lumbar spine (LS) was higher in children who switched to an efavirenz (EFV)-based antiretroviral therapy regimen from a ritonavir-boosted lopinavir (LPV/r)-based regimen in comparison to those who remained on a LPV/r-based regimen (-0.68 vs -1.20, p<0.01 and 0.01 vs -0.45, p<0.01, respectively). Now, we assess whether improvements in bone accrual are durable after an additional 1 year follow-up. Methods: The CHANGES Bone Study follows HIV-infected children who participated in a randomized trial in Johannesburg, South Africa of pre-emptive switching to EFV (N=106) compared to remaining on LPV/r (N=113) for those initially suppressed on LPV/r. WB and LS BMC were assessed by DXA at baseline and 12 months (mean 2.1 and 3.1 years after ART switch, respectively). BMC Z-scores adjusted for sex, age, and height were generated using reference norms from the Bone Mineral Density in Childhood Study. We compared LPV/r and EFV groups by intent-to-treat. Results: A total of 214 children (97%) completed both study visits, including 103 in the EFV group and 111 in the LPV/r group. The mean age was 7.4 years (SD 1.2). Similar to baseline, the WB and LS BMC Z-score at 12 months was significantly higher in those switched to EFV compared to those remaining on LPV/r (-0.77 vs. -1.26, p<0.01 and -0.07 vs. -0.53, p<0.01, respectively). The change in BMC Z-score from baseline to 12 months was not different between those switching to EFV and those remaining on LPV/r at the WB (-0.10 vs. -0.07, p=0.63) or LS (-0.08 vs. -0.08, p=0.99). However, WB and LS BMC Z-scores for both groups declined slightly relative to reference norms. At both baseline and 12 months, differences between the EFV and LPV/r groups were greater in the girls than the boys (Table 1). Conclusion: Bone mass remains higher in pre-pubertal perinatally-infected children 3 years after switch to EFV in comparison to children remaining on LPV/r. Differences between groups are greater in girls than boys. Rate of change in BMC did not differ between groups during the last year of observation, suggesting that benefits of switching to EFV for bone may be greatest close to the switch. These findings provide additional rationale for switching children with sustained viral suppression receiving first line regimen with LPV/r to EFV.

CROI 2017 354