1. Is Highly Active Antiretroviral Therapy (HAART) in pregnancy protective against maternal mortality? Results from a large DREAM cohort in Malawi and Mozambique Presented by Leonardo Palombi (Italy). G. Liotta 1, S. Mancinelli 1, E. Gennaro 2, P. Scarcella 1, K. Saines-Nielsen 3, N. Abdul Magid 4, P. Germano 5, J. Haswell 6, G. Guidotti 7, E.Buonomo 1, A. Doro Altan 1, L. Palombi 1, MC. Marazzi 8 1 University of Tor Vergata, Public Health Dept., Rome. Italy. 2 University “G. D’Annunzio”, Chieti, Italy 3 UCLA. Pediatrics. Los Angeles, USA 4 DREAM Program. Community of S. Egidio, Maputo, Mozambique. 5 DREAM Program. Community of S. Egidio, Rome, Italy. 6 DREAM Program. Community of S. Egidio, Blantyre, Malawi. 7 Pisa University, Italy 8 LUMSA University, Rome, Italy TUAB0201

2. Background  HIV-1 infection is a major cause of maternal mortality (MM) in resource-limited settings.  The DREAM (Drug Resource Enhancement against AIDS and Malnutrition) Program routinely offers triple ART from pregnancy throughout the postpartum period regardless of maternal CD4 cell count for prevention of mother to child HIV transmission (PMTCT) at clinical centers in sub-saharan Africa.  We evaluated the uptake of triple ART and maternal mortality in our PMTCT cohort at our centers in Malawi and Mozambique.

3. In prior cohort studies, our group reported a significant association between duration of triple ART in pregnancy and maternal mortality

4. Methods  Study design: Retrospective cohort  Records for all HIV+ pregnant women accessing 16 DREAM centers in Malawi and Mozambique from 6/2002 to 6/2010 were reviewed. Group 1: Women starting triple ART during pre- natal care, n = 8169. Group 2: Pregnant women on established triple ART for their own health, n = 1981  Women identified in pregnancy were offered NVP- based triple ART at 14 weeks (if required for health) or 25 weeks gestation until 6 months for PMTCT.

5. Specific Aims 1.To determine risk factors associated with maternal mortality during pregnancy until 6 weeks postpartum. 2.To compare mortality, ART uptake, loss to follow- up and laboratory parameters between women receiving triple ART for PMTCT and women on established ART during pregnancy. 3.To evaluate mortality in women beyond the immediate postpartum period according to duration of triple ART exposure with adjustment for time in medical care and duration of antenatal ART.

6. Baseline parameters: 10,150 women from 6/02 – 6/10 in Malawi + Mozambique MedianIQR: 25 - 75 Age26 years 23-30 CD4 cells/ mm3392258- 563 Log HIV-13.93.2- 4.4 BMI23.421.5 – 25.7 Hemoglobin10.09.0 – 11.0

7. Comparison between groups 7 Baseline parametersGroup 1 N=8,169 Mean (SD) Group 2 N =1,982 Mean (SD) p HAART in days832 (586) 1459 (621) <0.001 CD4 cell count431 (275) 354 (366) <0.001 Viral load cps/ml3.6 (1.3) 3.8 (1.5) <0.001 Body Mass Index23.8 (3.6) 22.0 (3.8) <0.001 Hemoglobin9.9 (3.4) 10.1 (1.9) 0.26 Refusal/Loss to follow up 9.83.9<0.001 Maternal Mortality1.10.70.17

8. Maternal mortality (MM)  Deaths up to 42 days postpartum: 101 (0.99%) Group 1 (ART during PMTCT): 87/ 8169 (1.1%) Group 2 (established ART): 14/1981 (0.7%) Deaths by CD4 cell count strata: < 350 cells/ mm 3 : 1.3% > 350 cells/ mm 3 : 0.7% Deaths by triple ART exposure: < 30 days of antenatal ART: 22/ 991 (2.2%) > 90 days of antenatal ART: 35/ 5260 (0.7%) Mozambique MMR 520 (0.52%) Unicef 2009 Malawi MMR 1100 (1.1%) Unicef 2009

9. Multivariate analysis of Maternal Mortality: logistic regression forward stepwise model Outcome variable MaternalDeath OR95% CLp Shorter ART exposure1.721.36 – 2.170.001 Baseline Viral Load0.840.61 – 1.170.3 Baseline CD4 count1.571.19 – 2.06< 0.001 Baseline Hemoglobin1.421.06– 1.910.02 Baseline BMI1.981.47 – 2.64< 0.001

10. Maternal Mortality according to pre-delivery duration of triple ART 10 Days of antenatal triple ART Antenatal observation days Mean (CL95%) No. of deaths%N 0 - 30154 (124-185)222.2991 30 - 90175 (166-185)441.13899 91 - 270262 (251-274)210.63282 > 2701059 (733-1386)*140.71978 Total373 (309-438)1011.010150 * p<0.001

11. Mortality in women over 4 years adjusted for observation time under triple ART Days of antenatal triple ART Total Observation time - days mean (CL95%) No. of deaths%N 0-30779 (738-821)454.5991 31-90817 (799-835)922.43899 91-270865 (846-885)581.83282 > 2701459 (1431-1486)392.01978 Total954 (942-967)2342.310150

13. Risk factors for mortality up to 4 years following delivery- Cox regression analysis 13 Sig.HR95.0% CI for HR LowerUpper HAART >270 (reference) 0.0001.0 0-30 days 0.0005.53.58.7 31-90 0.0002.81.94.3 91-270 0.0031.91.22.8 Viral Load > 5Log (reference) 0.0081.0 <4 C Log /ml 0.0250.60.30.9 4.0-5.0 0.8570.90.61.4 CD4 >350 0.0001.0 < 200 Cell/microL 0.0001.91.32.7 200-350 0.0011.71.22.3

14. Sig.HR95.0% CI for HR LowerUpper BMI >20.0 (reference) 0.0001.0 <18.5 0.0002.71.84.0 18.5-20.0 0.0121.71.12.5 Hb >10.0 (reference) 0.0191.0 8.0-10.0 gr/100cc 0.0171.61.12.4 <8.0 0.0161.41.11.9 Risk factors for mortality up to 4 years following delivery- Cox regression analysis (2)

15. CD4 Viral Load

16. Hemoglobin BMI

17. Summary  Maternal mortality is significantly associated with duration of triple ART, baseline CD4 cell count, BMI and degree of anemia.  Women on established ART while pregnant, women with CD4 counts > 350 and women with > 90 days of antenatal ART have a lower mortality risk during pregnancy and immediate postpartum period.  Women with extended antenatal ART (beyond 270 days prior to delivery) have the highest four year survival benefit.

18. Summary  Provision of triple ART in pregnancy carries a significant survival benefit and renders maternal mortality in HIV+ women (1%) equivalent or lower to population country- wide statistics and significantly lower than the mortality of untreated HIV+ mothers in Malawi (6%). Lettow 2011  Women who present to medical care late in pregnancy and receive shorter courses of triple ART (lowest triple ART exposure in absolute terms) have a higher mortality risk over 4 years. 18

19. Conclusions  Triple ART for PMTCT purposes have a significant impact on maternal mortality in resource-limited settings.  Extended triple ART carries a survival benefit extensive to all women regardless of CD4 cell strata.  Provision of triple ART to all HIV-infected women significantly decreases the risk of maternal death in the short and medium term. 19

20. Grazie! To our patients To our sponsors To the members of the DREAM Program