Introduction

With the implementation of recommendations for universal prenatal HIV counselling and testing, the administration of combination antiretroviral therapy (cART), scheduled caesarean section (CS) delivery and avoidance of breastfeeding, mother-to-child transmission (MTCT) rates dropped from around 20–25 to 1–2 % in the developed country settings [9]. In 1999, the results of the unique randomized controlled trial of vaginal delivery vs elective CS proved an 80 % efficacy for planned elective CS, while a large international individual patient meta-analysis reported a 50 % decreased MTCT risk associated with elective CS [7]. The administration of antiretroviral drugs in pregnancy, initially zidovudine (ZDV) as a monotherapy and subsequently cART, was the key factor behind reduced MTCT rate [9]. Several studies and in particular the study from the European collaborative team showed that maternal viral load was the preeminent independent risk factor for MTCT, with a viral load >3 log10 copies/mL associated with 12-fold increased risk of transmission [5].

Case

Here we report the clinical, virological and immunological outcome of the three pregnancies of a woman infected by HIV through sexual contact. The woman was diagnosed to be HIV infected in 1998, her CDC classification was A2 and her CD4+ cell count was 736 cells/mm3, the virus was B subtype and the viral load was <2.7 log10 copies/mL. She became pregnant for the first time on 1999, administered with ZDV as prophylaxis during gestation and intrapartum and delivered a daughter by elective CS at 38 weeks of gestation, on October 2009. The baby was administered ZDV until 6 weeks of age and was not breastfed. She was regularly followed up by our clinic and was HIV uninfected.

The second pregnancy was in 2000 and the same protocol was applied to the mother and the newborn. The baby, born on December 2000, was a daughter, had regular follow-up at our clinic and was HIV uninfected as well.

In May 2004, the woman’s CD4+ cell count was 321 cells/mm3 and had a viral load of 5 log10 copies/mL, and she started a cART holding ZDV + lamivudine (3TC) + lopinavir/r (LPV/r). This therapeutic approach leads to an undetectable viral load within 2 months. On August 2005, her viral load was still undetectable; however, the cART was simplified in tenofovir (TDF) + emcitrabine (FTC) plus efavirenz (EFV), administered in a single tablet in order to support adherence.

In April 2010, she became pregnant, for the third time, unplanned. Her viral load was undetectable. The gestation was confirmed at 7 weeks; EFV was immediately interrupted, and after 2 weeks, TDF and FTC were interrupted as well and the cART therapy was replaced by ZDV + 3TC + LPV/r. During gestation, she complied with the therapy and the therapy was well tolerated. Viral load, checked five times, was undetectable and at the end of the first quarter, during the second quarter and in the middle of the third quarter (the last undetectable viral load dates back to the 35th weeks of gestation). CD4+ cell count was checked along with the viral load: 544 cells/mm3 at the end of the first quarter, 823 cells/mm3 during the second quarter and 671 cells/mm3 in the middle of the third quarter. During the last 3 weeks of gestation, the mother suffered from gastroenteritis with nausea and vomiting, lasting for about 10 days. Thus, she was unable to take the cART regularly and she missed about half the doses she was supposed to take. What is more, she did not inform her physician about the symptoms and she did not take any kind of treatment. She delivered a male newborn at 38 weeks of gestation by an elective CS, ZDV was administered during the intrapartum period and the newborn received ZDV for 6 weeks and was not breastfed. At delivery, the mother’s viral load was 5.8 log10 copies/mL and 671 cells/mm3 for the CD4+ cell count. Twenty days after the delivery, the mother resumed the cART regimen including TDF, EFV and 3TC and she was reported to have an undetectable viral load 1 month later.

During the first 3 months of life, the baby showed a negative HIV-DNA in four tests (the first test was performed at delivery and later at 1, 2 and 3 months of age) with CD4+ cell count ranging from 1,672 to 3,055 cells/mm3. At 4 months of age, his HIV-DNA was positive and HIV-RNA load was >7 log10 copies/mL; these data were confirmed at 5 months of age. He was classified as A1 according to CDC pediatric classification and moved to a cART therapy with ZDV + 3TC + LPV/r that led to an undetectable viral load (<37 copies/mL) 3 months after.

Discussion

Two important features are highlighted in this case. One is the risk of MTCT when antiretroviral treatment is interrupted during pregnancy. Galli et al. recently showed that the rate of MTCT is 4.9 % (95 % CI, 1.9–13.2 %) when cART is interrupted in the first quarter, while it is much higher [18 % (95 % CI, 4.5–72.7 %)] when cART is interrupted in the third quarter [6]. Unfortunately, the interruption of cART in our case occurred in the third quarter, in particular in the last 3 weeks of gestation. It is possible that even this short period of interruption led to a peripartum vertical HIV transmission. Thus, the time course of positive HIV-RNA and DNA diagnostic studies observed in the baby clearly supported an acquired intrapartum HIV infection. According to Bryson et al., in utero HIV infection transmission should be diagnosed only when the diagnostic study becomes positive during the first week of life [1]. Secondly, this case showed the risk of MTCT when higher viral load occurs near the time of delivery. These data are confirmed by several authors and a recent paper indicates that the viral load >4.78 log10 copies/ml is independently associated with an increased rate of MTCT [1, 4]. This specific viral load on the day of delivery was very high (5.8 log10 copies/ml).

Lastly, an adequate cART compliance level is a challenge in pregnancy due to health factors such as advanced AIDS stage and health-related symptoms (nausea, vomiting, fatigue) and the possible presence of comorbidity. Our patient experienced gastroenteritis and the mother was not able to comply with the prescribed cART protocol and with the drug exposure. Unfortunately, therapeutic drug monitoring result was not available. However, some indirect evidences are available to support our assumptions that a suboptimal cART exposure may increase an MTCT of HIV infection. Firstly, it was recently proved that diarrhoea was associated with a reduced exposure to cART [3]. Secondly, growing number of evidences are available showing that plasma cART exposure influences antiviral drug activity [8]. Accordingly, it was recently suggested that suboptimal cART exposure during pregnancy can lead to HIV RNA rebound, the selection of resistant virus or an increased risk of HIV-1 transmission to the infant [2].

In conclusion, our case proved the importance of compliance with cART during pregnancy and that is crucial for pregnant women to disclose new health issues (e.g. gastroenteritis) to allow the physician to adapt prophylactic strategies [10]. Perhaps, CS scheduled before 38 weeks of gestation and the administration of combination neonatal prophylaxis rather than zidovudine monotherapy could have helped avoiding this MTCT of HIV infection.