Laboratory indicators of mastitis do not predict the risk of mother-to-child transmission of HIV in Zimbabwean women

Current laboratory indicators of mastitis are not useful predictors of HIV levels in breast milk, according to findings of a cross-sectional study published in the August 15^th edition of the Journal of Infectious Diseases. According to the study, milk cell counts and milk electrolyte concentrations can not be used to assess the risk of mother-to-child transmission (MTCT) of HIV.

About 1,000 children become infected daily with HIV by vertical transmission and 500 of these are associated with breastfeeding. Mothers in developing countries practise either exclusive breastfeeding, breastfeeding plus formula feeding (mixed feeding), or formula feeding alone, depending on their socioeconomic status.

There are conflicting reports about the risks of mother-to-child transmission of HIV which are associated each of these feeding practices. Some studies have reported that exclusive breastfeeding reduces the rate of MTCT while others report that breastfeeding increases MTCT. Indeed, HIV-1 RNA and HIV-1 DNA in breast milk and mastitis are associated with the risk of MTCT through breast milk.

A greater understanding of the relationship between indicators of inflammation and HIV loads in breast milk could improve mother-to-child transmission prevention strategies targeting high-risk breastfeeding. These interventions would depend on the ability to identify women with high HIV milk levels due to mastitis using simple inexpensive laboratory assays suitable for resource-poor settings.

In an effort to identify indicators which predict HIV shedding in breast milk, a team of US and Zimbabwean investigators evaluated the relationship between laboratory indicators of mastitis and HIV-1 RNA and DNA loads in the breast milk of Zimbabwean women.

The study took place in clinics for HIV-infected mothers and infants in Chitungwiza and Epworth, two towns in Zimbabwe. The participants were HIV-positive lactating women 6-16 weeks after delivery who consented to take part in the cross-sectional study. Breast milk and blood samples were collected from each participant and a standard questionnaire was filled regarding current symptoms of breast inflammation and uptake of antiretroviral treatment (ART).

In all, 217 breastfeeding women were enrolled and 414 samples of breast milk were obtained from 216 of the 217 participants. None of the women were on ART but 85% reported having received single dose nevirapine for the prevention of MTCT.

Milk was processed to obtain lactoserum and cell pellets for total and differential white blood cell counts using standard microscopic examination procedures. Sodium (Na+) and potassium ion (K+) concentrations in lactoserum were determined for each woman. Asymptomatic mastitis was defined as either Na+ concentration > 12 mmol/L, Na+/K+> 1, or total leucocyte count > 106 cells/mL of milk.

HIV-1 RNA in lactoserum was measured using the AMPLICOR kit whereas HIV-1 DNA was determined by real-time polymerase chain reaction using DNA extracted from whole breast milk.

Symptomatic mastitis was reported at enrollment in 8% of 217 women and 5% of 434 breasts. Asymptomatic mastitis as defined by total milk cell counts or electrolyte levels was detected in 23-28% of breast milk samples. Na+ concentration > 12 mmol/L and Na+:K+ >1 as indicators for asymptomatic mastitis were in agreement in the majority of cases but were discordant for 3% of the samples. Similarly, while Na+ concentration > 12 mmol/L and total leucocyte count > 106 cells/mL were in agreement they were discordant for 19% of the samples.

HIV-1 RNA was detected in 239 (58%) of 409 breast milk samples from 142 (66%) of 215 women. Although the median HIV-1 RNA levels in breast milk was significantly lower than that in plasma, viral levels in milk and plasma were associated (P

HIV-1 DNA was detectable in 85% of 201 breast milk cell pellets but was not associated with the indicators of mastitis. There was a significant correlation between the detection of HIV-1 DNA and HIV-1 RNA in milk samples.

The usefulness of breast milk Na+ concentrations, Na+:K+, and milk neutrophil counts as predictors of milk HIV-1 RNA and DNA loads were evaluated based on previously published cut-off points which have been associated with an increased risk of MTCT. These laboratory indicators of mastitis were not able to predict breast milk HIV-1 RNA loads with the desired sensitivity and specificity required for a diagnostic assay.

In conclusion, neither milk cell counts nor electrolyte concentrations were useful predictors of milk HIV-1 RNA or DNA loads for individual Zimbabwean women.

Additional studies are urgently required to increase our understanding of the factors which increase HIV-1 DNA and HIV-1 RNA in breast milk of HIV-positive mothers. These studies might provide invaluable insights for the development of low cost assays which can identify high-risk breastfeeding mothers as part of MTCT control efforts.