Pharmacokinetics of antiretrovirals in pregnancy

CROI 2014Polly Clayden, HIV i-Base

Physiological changes during pregnancy can influence the pharmacokinetics (PK) of antiretrovirals and decrease drug exposure. Opinions on whether this is of clinical importance and recommendations for dosing differ.

Four posters at CROI 2014 reported PK data for atazanavir, darunavir and raltegravir suggesting that routine dose adjustment in pregnancy is unnecessary. The PANNA Network of European centres – set up to collect PK data in pregnant women in Europe – conducted three of these studies. [1] David Burger and investigators at the Radboud University Nijmegen Medical Centre, the Netherlands lead this network.


Ritonavir-boosted atazanavir (ATV/r) is a preferred protease inhibitor for antiretroviral (ARV)-naive pregnant women in European (Including BHIVA) and US perinatal guidelines. The package insert recommends a dose increase in the second and third trimesters of pregnancy with concomitant TDF use. Conflicting data exist about the influence of TDF on ATV concentrations, and whether or not this dose adjustment is necessary.

Results were presented from a French multicentre, cross sectional cohort study investigating steady state ATV plasma concentrations at 24 hours post-dose (C24h) in HIV positive pregnant women. [2]

Women receiving an ATV/r 300/100 mg once-daily containing regimen with available viral load, CD4 and demographic data were enrolled. None of the women in this study received an adjusted dose of ATV.

ATV C24h was determined at delivery, during each trimester and post-partum using UPLC-MA/MS; the target was 0.15 mg/L.

Virological failure was defined as two consecutive viral load results >50 copies/mL within two months prior to delivery.

The investigators also collected weight, gestational age, and APGAR score data from neonates.

A total of 103 pregnant women were evaluated. The women were a median of: 34 years (IQR 31to 37), 7 years (3 to 9) since HIV diagnosis, 5 years (2 to 8) on ART and 1 year (0 to 1) of ATV/r; 13% started treatment in pregnancy. Almost half the women (42%) received a TDF/FTC backbone

A minority: 6%, 4% and 2% were coinfected with HBV, HCV and HBV/HCV, respectively. BMI before pregnancy was a median of 25 (IQR 21 to 28); baseline CD4 197 cells/mm3 (IQR 101 to 290) and viral load 46,042 copies/mL (IQR 31,066 to 102,800). The majority of women (88%) were from sub-Saharan Africa.

A total of 366 samples were available for analysis. The investigators reported statistically similar ATV C24h values for the different trimesters and delivery compared to post partum: p=0.11 for 1st trimester, p=0.20 for second trimester, p=0.69 for third trimester and p= 0.76 for delivery. The majority of women (85%) were above the 0.15 mg/L threshold across all trimesters. Coadministration of TDF did not affect ATV C24.

Cord plasma ATV concentration was a median of 0.146 mg/L (IQR 0.64 to 0.295) in a subset of 28 maternal/infant pairs. Cord/maternal plasma ATV ratio was 0.19 (0.10 to 0.32) consistent with low transplacental transfer and the protein binding ratio of ATV.

There was no significant change in CD4 count from the start of ATV/r containing treatment to post partum. At delivery, 97% of women were virologically suppressed.

There were no vertical transmissions in this cohort. Of 88 women accessed, 40 delivered vaginally and 48 by caesarean section. Of 82 neonates with available data, 16 were preterm (<37 weeks gestation).

Infant characteristics were median: gestational age 38 weeks (IQR 37 to 39) weeks, weight 2,970 g (IQR 2,755 to 3,358) and APGAR score 10.

Only one woman presented during delivery with grade 3 hyperbilirubinaemia. Of 48 infants, 24 presented with grade 2 and 5 with grade 3 hyperbilirubinaemia.

A related study from the PANNA Network looked at steady-state intensive 24-hour PK in the third trimester and at least two weeks postpartum among pregnant women receiving 300/100 mg ATV/r in regimens with and without TDF. [3]

ATV and RTV plasma concentrations were determined by validated UPLC method.

The investigators calculated geometric mean ratios (GMR) and 90% confidence intervals for PK parameters third trimester/postpartum and compared those with and without TDF.

This analysis included 29 women of whom 11 were treatment naive at conception; 15 of the women were black and 14 white; 19/29 received concomitant TDF. Approaching delivery 76% were virologically suppressed.

Paired PK curves (third trimester and postpartum) were available for 25 women. Exposure to ATV in the third trimester was 34% lower than post partum.

GMR (90% CI) of ATV PK parameters third trimester/postpartum overall were: 0.66 (0.57 to 0.75) for AUC0-24h; 0.70 (0.61-0.80) for Cmax ; 0.59 (0.48 to0.72) for C24h, all comparisons p<00.1.

No statistical difference in AUC24h was found between women receiving TDF vs no TDF: GM (95%CI) third trimester 28.8 mg.h/L (22.2 to 37.4) vs 32.08 mg.h/L (21.1 to 48.7); postpartum 46.1 (36.2 to 58.6) vs 49.2 mg.h/L (34.7 to 69.8).

None of the women had ATV concentrations <0.15 mg/L.

Viral load was detectable in six women around delivery. PK parameters for ATV were similar in these women to those with undetectable viral load around delivery.

The median cord blood/maternal plasma concentration ratio in 12 maternal/infant pairs was 0.20 (range 0.6 to 3.05).

Median gestational age at delivery was 39 weeks.

One infant died (congenital diaphragmatic hernia resulting in respiratory failure and septic shock). The investigators did not consider an association with ATV/r to be likely as it was started in week 21 of pregnancy in this case, and the closure of the pleuroperitoneal canal happens around week 8 of pregnancy.

There were no vertical HIV transmissions.

The investigators noted that therapeutic drug monitoring of ATV should be considered for treatment experienced pregnant women with PI resistance.


European and US perinatal guidelines include ritonavir-boosted darunavir (DRV/r) as an alternative protease inhibitor in ARV-naive pregnant women.

The dosing recommendation for ARV-naive patients is DRV/r 800/100mg once daily but PK data in pregnant women are limited.

A PANNA evaluation of this dose DRV/r during pregnancy performed intensive steady-state 24-hour PK profiles in the third trimester and at least two weeks postpartum, as in the ATV study described above. [4]

Target concentration for DRV is 0.55 mg/L (EC50 for resistant virus).

Of 15 women included in the analysis, 7 were black and 8 white. A third of the women were treatment naïve at conception; 2 conceived on DRV/r; 4, 8 and 1 started DRV/r in the first, second and third trimesters respectively. Seven paired PK curves (3rd trimester and postpartum) were available.

GMR (90% CI) of DRV PK parameters third trimester/ postpartum were: 0.63 (0.51 to 0.77) for AUC0-24h; 0.72 (0.57 to 0.93) for Cmax; 0.36 (0.22 to 0.58) for C24h.

The mean DRV free fraction (95% CI) was 12% (11 to 14%) in the third trimester and 10% (7 to 13%) postpartum. Two of 15 women had DRV concentrations <0.55 mg/L in the third trimester, vs none postpartum.

The median cord blood/maternal ratios in 6 maternal/infant pairs were 0.12 (range 0.08-0.35).

The median gestational age at delivery was 38 weeks. Around delivery 4 women had viral load >50 copies/mL. Two weeks prior to delivery one woman still had a significant viral load (28,711 copies/mL). She was thought to be non-adherent and had directly observed therapy until delivery.

There were no vertical HIV transmissions.

The investigators noted that contrary to previously published data for DRV exposure with DRV/r 600/100 mg twice daily, the decrease in exposure with 800/100 mg once daily does not appear to be compensated by a higher free fraction during pregnancy.

They suggested DRV/r 600/100mg BID to be the preferred dose during pregnancy for treatment-experienced women.


According to European and US perinatal guidelines, raltegravir (RAL) can be used in HIV positive pregnant women in special circumstances, because safety and PK information is limited.

RAL has been shown to rapidly reduce viral load. BHIVA guidelines recommend RAL for late presenters (>28 weeks) and women with unsuppressed viral load close to delivery.

In the third PANNA evaluation, women receiving RAL 400 mg twice daily during pregnancy also had intensive steady-state12-hour PK profiles in the third trimester and at least 2 weeks postpartum. [5]

The suggested target for RAL is 0.02 mg/L (levels below this threshold were associated with viral failure in QDMRK).

This analysis included 14 women of whom 8 were black and 6 white. Five women were treatment naive at conception, 3 conceived on RAL and 2, 4 and 5 started in the first, second and third trimesters respectively. RAL was combined with a PI-based regimen in 9/14 women; 6/14 women received quadruple ART regimens.

Paired PK curves (third trimester and postpartum) were available for 12 and third trimester only for two women.

Geometric mean ratios (90% CI) of RAL PK parameters third trimester/postpartum were: 0.77 (0.59 to 1.00) for AUC0-12; 0.83 (0.55 to1.25) for C max; 0.54 (0.28-1.05) for C12h; concentrations in the 3rd trimester were respectively: 4.95 (3.01 to 8.13) mg.h/L, 1.40 (0.74 to 2.65) mg/L and 0.05 mg/L (0.03 to 0.09) mg/L.

One woman had a C12h third trimester level below the suggested threshold of 0.020 mg/L. Her viral load was 74 copies/mL in the 3rd trimester and undetectable at delivery.

The median ratio of cord blood/maternal RAL concentrations in 8 mother/infant pairs was 1.24 (range 0.13-4.53).

Median gestational age at delivery was 38 weeks and birth weight was 3115 g. Viral load was detectable in 3 women around delivery; all were <400 copies/mL and none of their levels were below the suggested threshold.

None of the infants were HIV infected and no birth defects were reported.

The investigators reported that the decrease observed in exposure to RAL during 3rd trimester compared to postpartum: 33% for AUC0-12 and nearly 50% for C12h, is not considered to be of clinical importance. There is considerable interpatient variability in PK reported with RAL, which was also demonstrated in this study. RAL showed good transplacental transfer.


An important component of ensuring the safe and effective use of antiretrovirals during pregnancy is studying their PK.

To date, most protease inhibitors have shown reduced total plasma concentrations during the third trimester. But pregnancy-related changes in the binding of protease inhibitors also need to be considered. In some cases this partially compensates for the reduction in the total concentration by increasing the protein-free bio-available fraction.

Whether reductions in plasma concentration should lead to dose-adjustment during pregnancy is not always clear-cut. This has certainly been the case for lopinavir/ritonavir in the past and is again with boosted atazanavir and darunavir. The importance of any reduction in total or protein-free drug must also factor in the likelihood of achieving therapeutic efficacy – this can be guided by the absolute concentrations compared to the target concentrations – which in both studies were adequate for atazanavir and then by the clinical outcome, which should include both early and long term viral suppression and of course transmission data.

Such data are emerging for atazanavir and are needed for darunavir. In the latest version – the 2014 interim review – the BHIVA guidelines have taken a cautious but pragmatic view, recognising that twice daily might be preferable to once daily darunavir on the basis of the PK studies but that for many patients established on a fully effective once daily regimen the case for switching to twice daily may not outweigh the advantages of continuing with once daily.

Conversely, where there is less certainty of effect, initiating ART during pregnancy or where HIV is known or suspected to harbour drug resistance mutations, twice daily for the duration of pregnancy should be considered until more clinical and virological data are available.


Unless otherwise stated all references are to the programme and abstracts from the 21st Conference on Retroviruses and Opportunistic Infections (CROI), 3-6 March 2014, Boston.

  1. PANNA Network.
  2. Lê MP et al. Safety, efficacy, and PK of atazanavir/ritonavir (300/100 mg QD) in HIV+ pregnant women cohort. Poster abstract 889. (PDF)
  3. Colbers A et al. Effective exposure to atazanavir during pregnancy, regardless of tenofovir use. Poster abstract 892. (PDF)
  4. Colbers A et al. Low darunavir exposure during pregnancy with 800/100 mg darunavir/r QD dosing. Poster abstract 887. (PDF)
  5. Blonk M et al. A comparison of the pharmacokinetics of raltegravir during pregnancy and postpartum. Poster abstract 890. (PDF)

Links to other websites are current at date of posting but not maintained.