Easier Said Than Done: World Health Organization Recommendations for Prevention of Mother-to-Child Transmission of HIV—Areas of Concern

E ditor: The World Health Organization (WHO) released recommendations on treatment, prevention, and infant feeding practices within the context of HIV infection based on the “latest scientific evidence” available. ¹ The “Rapid Advice” document anticipates the release of official HIV Prevention-of-Mother-to-Child Transmission (PMTCT) guidelines. Public health programs providing HIV care in sub-Saharan Africa such as DREAM (Drug Resource Enhancement against AIDS and Malnutrition) are concerned about the vast ramifications following implementation of specific recommendations.

DREAM provides HIV care to 85,000 individuals in 10 African countries and was the first organization to introduce triple antiretroviral therapy (ART) for HIV PMTCT purposes regardless of maternal immunologic or virologic conditions. In the program, 10,000 pregnancies have been followed with encouraging results. ² Although the recommendations undoubtedly underscore significant advancements in HIV PMTCT, based on our experience, there are areas of concern.

First, the requirement for “CD4 cell counts to be available antenatally so that decisions can be made in regards to maternal ART eligibility” can become a double-edged sword. Although treating all patients with CD4 counts <350 cells/mm³ is a major advancement, the reality is that in most of sub-Saharan Africa, CD4 cell count measurements are not routinely available. The need for CD4 cell measurements before ART initiation will certainly delay treatment initiation in pregnancy. It is critical that guidance be provided to ensure that ART initiation not be postponed. A caveat for provision of ART regardless of CD4 cell counts, under these circumstances, should be included. As currently written, triple ART for patients with WHO clinical stages 1 or 2 without CD4 cell count results is not recommended, even though a significant proportion (>50%) of patients in these stages actually have CD4 cells under 350/mm³. ³

The recommendations are for initiation of zidovudine monotherapy (option A). As CD4 cell monitoring is not guaranteed, patients could inadvertently be prescribed prolonged courses of zidovudine alone, a practice long abandoned in developed countries. Exceptions should be envisioned for populations without access to flow cytometry.

The recommendations also leave space for massive use of efavirenz (EFV) during pregnancy including the first trimester. Although efavirenz is widely available in public health care programs in sub-Saharan Africa, there is an extreme paucity of data regarding its use in early pregnancy. In developed countries, EFV use is restricted during early gestation. FDA recommendations state: “Sustiva may cause fetal harm when administered during the first trimester to a pregnant woman, pregnancy should be avoided in women receiving Sustiva” (FDA category D and contraindicated during breastfeeding). The AMATA study provided limited information on the use of EFV in the last trimester of pregnancy and during breastfeeding. ⁴ In that study, all breastfed infants had detectable levels of EFV in plasma (unknown toxicity potential). These recommendations imply life-long use of EFV for women with lower CD4 cell counts. This can lead to a large number of conceptions by women on EFV. In resource-limited settings, knowledge of gestational age is limited and pregnancy outcomes are not generally monitored. The recommendation to use a potentially teratogenic drug wide scale in a vulnerable, disenfranchised population is questionable, considering that there are alternate ART available. In addition, EFV can potentially carry the same risk of hepatic toxicity in pregnancy as nevirapine.

Finally, the WHO document recommends daily administration of nevirapine (NVP) to HIV-exposed infants throughout breastfeeding. The safety and efficacy of extended use of NVP to infants beyond the first few weeks of life are not yet established as data from large-scale clinical studies are not available. Studies such as SWEN and PEPI trials explored 6 or 14 weeks of NVP administration to infants. The potential risks of NVP in women with CD4 >250/mm³ and men with CD4 >400/mm³ have restricted its use. There is an asymmetry in contraindicating NVP use in HIV⁺ patients with higher CD4 cell counts but not in HIV-uninfected immune-competent children. Moreover, studies demonstrate significant plasma levels of NVP in breastfed children of mothers receiving NVP-based ART. ⁵ Double exposure of infants to maternal NVP and directly administered NVP needs to be addressed as potential toxicity may occur. In conclusion, WHO guidelines are a significant improvement to earlier versions, but caution should be exerted in the interpretation of specific recommendations, particularly in the absence of supportive data.