The use of prophylaxis against HIV after sexual exposure in pregnancy to reduce the risk of vertical transmission

Case Report

A 27-year-old female with a twin-pregnancy was tested negative for HIV at routine antenatal screening. At 27 weeks of gestation she attended our service with her new partner of two months. For the previous two weeks they had several episodes of unprotected anal (UPARI) and vaginal intercourse (UPVRI). Her partner was tested HIV-positive with Abbott Laboratories-Determine™ HIV-1/2 finger prick rapid test (Abbott Laboratories, Maidenhead, Berks, UK). Serological testing confirmed his diagnosis of HIV-1 infection; (viral load [VL] of 5230 copies/mL and CD4 630 cells × 10⁶/L). HIV antibody tests on the patient were negative including a finger prick rapid test and Bayer ADVIA Centaur (HIV-1, HIV-2 and HIV-1 subtype O). HIV RNA was also not detected (<50 copies/mL) (Roche COBAS Amplicor; Roche Diagnostics GmbH). The couple was advised to use condoms consistently or abstain from sex. Five days later, after discussion with the couple and colleagues, antiretroviral therapy (ART) was started, in order to minimize the risk of vertical transmission in the event of maternal seroconversion, using zidovudine, lamivudine and lopinavir boosted with ritonavir.

As she was in the window period, further reference laboratory tests were performed (Centre for Infections, Colindale, London, UK). HIV-1 proviral DNA was not detected. HIV-1 p24 antigen (p24Ag) testing (Biorad enzyme immunoassay [EIA]) was negative, as was a combined HIV-1/2 antibody/p24Ag test (Integral Ag/Ab EIA). However, a second combined HIV-1/2 Ab/p24Ag test (Abbott Murex combined EIA) was reactive on the same sample. HIV testing was therefore inconclusive, although the reactivity in the Murex Ag/Ab assay was felt to be non-specific. Fortnightly samples for HIV proviral DNA, p24Ag and antibody testing remained unchanged throughout the course of her pregnancy with only the Murex Ag/Ab assay remaining reactive. By 31 weeks of gestation she was admitted under the care of the obstetricians because of problems with umbilical flow, and an elective caesarean section was performed at 36 weeks of gestation. As she was still in the window period for HIV-antibody testing based on her last episode of unprotected sex, IV zidovudine was given during delivery after which ART was stopped. She was advised that she might breastfeed due to concerns regarding the wellbeing of the premature infants. ART was not given to the infants.

Discussion

The management outlined above was based on an assessment of the maternal risk of acquiring HIV and the potential effects of seroconversion in pregnancy. The risk of acquiring HIV through a single episode of UPARI with an HIV-positive partner is 0.1-3.0% and 0.1-0.2% for UPVRI. ¹ Our patient had multiple exposures over a two-week period. The VL is the chief predictor of the risk of heterosexual transmission of HIV-1. ² One study of serodiscordant couples found no instance of heterosexual transmission among subjects with a serum viral load (VL) of <1500 copies/mL. They also found a dose-response relation of increased transmission with increasing VL, each log increment in the VL was associated with a rate ratio of 2.45 for seroconversion. ² The partner's VL in the case presented was 5230 copies/mL.

One study had found that women were more likely to acquire HIV during pregnancy (adjusted incidence ratio: 2.16 vs. 1.1 per 100 person-years), which might be attributable to hormonal changes affecting the genital tract mucosa. ³

These factors may have increased our patient's risk of acquisition to over 0.1-3.0%.

It has been hypothesized that mother-to-child HIV transmission (MTCT) may be higher in women who seroconvert during pregnancy than for those who are seropositive before pregnancy as the concentration of HIV in blood is high between the weeks of infection and seroconversion. ⁴ This was not supported by a prospective study which found no increased risk for transmission between women who seroconverted during pregnancy and those who were seropositive when first tested. ⁴ However, there have been cases reported of MTCT in women who seroconverted in pregnancy. ⁵

The occurrence of a twin pregnancy ⁶ and early delivery may also have increased the risk of MTCT. ⁷ The prospects of an early delivery (which was anticipated) raised concerns about our ability to suppress her VL in time for seroconversion to occur. Based on the potential consequences of HIV acquisition along with inconclusive testing we made the decision to monitor her closely for evidence of seroconversion and offer ART. Having ART on-board at the time of potential seroconversion might also help control her VL.

There are concerns regarding side-effects of ART for the mother and twins and also the potential effects on the pregnancy. Although combination therapy which includes a PI (as our regimen included) was associated with an increased risk of preterm delivery, ⁸ we felt that reports of nevirapine-associated hepatotoxicity ⁹ and animal studies suggesting a possibility for congenital abnormalities with the use of efavirenz, ¹⁰ limited our combination options.

Our patient and her twins have not acquired HIV. She decided not to breastfeed and the couple continued to use condoms.