CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Poster Abstracts

804 DOES CHANGING ART IN EARLY PREGNANCY FOR FETAL RISKS DESTABILIZE VIRAL SUPPRESSION?

805 COMPARISON OF FOUR CLASSICAL PI- AND RALTEGRAVIR-BASED REGIMENS DURING PREGNANCY

Violaine Peyronnet 1 , Josiane Warszawski 2 , Jeanne Sibiude 2 , Catherine Dollfus 3 , Christine Rouzioux 4 , Jérôme Le Chenadec 2 , Albert Faye 2 , Roland Tubiana 5 , Laurent Mandelbrot 2 1 Hôpital Louis-Mourier, Colombes, France, 2 INSERM, Le Kremlin-Bicetre, France, 3 Assistance Publique – Hôpitaux de Paris, Paris, France, 4 Paris Descartes University, Paris, France, 5 Paris Center Hospital Group, Paris, France Background: As increasing numbers of women with HIV are on becoming pregnant on antiretroviral therapy, and fewer infants are infected, there is concern about the possible risks for fetuses and infants exposed to ARV in utero. In case of potential toxicities or in the absence of good safety data, one option is to switch the maternal ART regimen as early as possible in pregnancy. Our aimwas to study the frequency and reasons of such treatment changes and to determine whether they led to poorer virological outcomes. Methods: All pregnancies in women with HIV1 enrolled in the national multicenter prospective French Perinatal cohort (EPF) were included if delivery occurred between 01/2005 and 12/2015 at 14 gestational weeks or more and the mother was on antiretroviral therapy (ART) at conception with a plasma viral load < 50 copies/ml. The reasons for switching the antiretroviral regimen (suppression or change of at least one drug) were analyzed in view of treatment guidelines at the time of the pregnancy, and defined as “for medical strategy” after excluding changes for side effects. Virological and pregnancy outcomes were studied by survival analysis censored at the time of ART switch and by logistic regression adjusted for a propensity score established for each patient according to baseline characteristics. Results: Of 10365 pregnancies with an outcome > 14 weeks’ gestation, nearly half (N = 4983) were on antiretroviral therapy at conception, and of these 1019 (9.8%) had a treatment switch in the first trimester of pregnancy. For 66.4% of these switches, the reason was a medical strategy to follow treatment guidelines. The proportion of switches was statistically higher when initial treatment was contraindicated (ORa: 21.9 [13.3-36.1]) or was regarded as an alternative (OR adjusted: 2.2 95% CI [1.3-3.7]), as compared to recommended first-line options. Treatment switches for medical strategy did not lead to poorer virological control, compared to pregnancies without such switches (19.3% vs. 16.2%, HRa: 1.0 [0.7-1.3]). Conclusion: Changing antiretroviral therapy early in pregnancy with the goal of improving fetal and pregnancy outcomes did not appear to have a destabilizing effect on viral suppression. However, the impact may differ according to the specific regimens and patient characteristics. Multidisciplinary preconceptional care must be provided to women with HIV, as well as further research on recent antiretroviral drugs to document their safety and risks.

Jeanne Sibiude 1 , Olivia Dialla 2 , Roland Tubiana 3 , Stephane Blanche 4 , Catherine Dollfus 5 , Pierre Frange 4 , Jérôme Le Chenadec 2 , Laurent Mandelbrot 1 , Josiane Warszawski 2 1 Hôpital Louis-Mourier, Colombes, France, 2 INSERM, Le Kremlin-Bicetre, France, 3 Pitié-Salpêtrière Hospital, Paris, France, 4 Necker Hospital, Paris, France, 5 Hôpital Armand-Trousseau, Paris, France Background: Antiretroviral therapy (ART) is recommended for all HIV-infected pregnant women as prevention of mother-to-child transmission. A large proportion of women now start pregnancy while already on ART, and more information is needed concerning tolerance of drugs received throughout pregnancy. We sought to compare 3 first-line recommended regimens in France (all of them including one ritonavir-boosted protease inhibitor and two NRTIs), and raltegravir-based ART which has become recently an option for pregnancy. Methods: All HIV-infected pregnant women included in the national French Perinatal Cohort between 2005 and 2015, receiving at conception a 3-drug regimen containing lopinavir/r, atazanavir/r, darunavir/r or raltegravir associated with 2 NRTIs (xTC+ ZDV, ABC or TDF). Immuno-virological, pregnancy and neonatal outcomes were compared, in intent-to-treat, according to treatment group using univariable and multivariable logistic regressions adjusting for NRTIs, maternal age, and geographical origin. Results: Among 3125 patients who were on ART at conception, we excluded 1017 patients with non-PI and non-integrase inhibitor based ART, 239 patients with 2-drug or 4-drug regimens and 272 patients with other drugs than those of interest. Overall, 1597 women responded to selection criteria, among which 45% (N=715) were receiving at conception a lopinavir-based regimen, 33% (N=536) atazanavir-based, 18% (N=288) a darunavir-based, and 4% (N=58) a raltegravir-based regimen. Change in ART during pregnancy was more frequent for atazanavir- and raltegravir- than for lopinavir-based regimens (AOR=1.4 [95%CI 1.0-1.8 and AOR=2.4 [1.4-4.3] respectively), but less frequent for darunavir-based regimens (AOR=0.6 [0.4-0.9]). In this population, only one child was HIV-infected. Immuno-virological outcomes, pregnancy and neonatal outcomes did not differ significantly among treatment groups (Table). Power was > 80% to detect a 2-fold increase in hospitalization, gestational diabetes, and preterm birth when comparing darunavir and atazanavir to lopinavir. Conclusion: Efficacy and tolerance during pregnancy were similar across the 3 groups of PI-based and raltegravir-based regimens. Notably, more recent drugs such as darunavir and raltegravir do not seem to be less safe than lopinavir and atazanavir which have been more largely described in pregnancy. Long term safety for children remains unknown.

CROI 2018 303