We analyzed breast milk samples collected from 51 Ugandan women 4 weeks after sdNVP administration. Genotyping results were obtained for all 10 breast milk samples with viral loads > 500 copies/ml using standard genotyping procedures (median: 1,208, range 503–8,509 copies/ml), and for 21 (51.2%) of 41 samples with viral loads <500 copies/ml using a nested PCR procedure (median: 25, range 25–200 copies/ml). Overall results were obtained for 31 (60.8%) of the 51 breast milk samples. HIV genotyping results were also obtained for plasma from these 31 women. Results from one woman were excluded because of a possible sample mix-up. The final data set used for analysis included results from 30 paired maternal plasma and breast milk samples (HIV subtypes: 15A, 1C, 12D, 2 intersubtype recombinant).
Twelve (40%) of 30 breast milk samples had at least one NVP resistance mutation; three of those samples had more than one mutation (). The most common mutations detected were K103N and Y181C. NVP resistance in breast milk was not associated with HIV transmission; NVP resistance was identified in breast milk from 4/11 (36.4%) transmitters vs. 8/19 (42.1%) non-transmitters (p=1.0). Among breast milk samples with NVP resistance, the median number of mutations detected was higher for the 11 women who transmitted HIV to their infants (2.0) than for the 19 women whose infants were HIV-uninfected (1.0); however, this difference was not statistically significant (p=0.18).
NVP resistance detected in plasma and breast milk samples (Uganda, 2003–2004)*.
Among the 30 women, five had NVP resistance mutations detected in plasma only, four had NVP resistance mutations detected in breast milk only, eight had NVP resistance mutations detected in both samples, and 13 did not have NVP resistance mutations detected in either sample. Five of the eight women who had NVP resistance mutations detected in both plasma and breast milk had different mutations detected in the two samples ().
We examined persistence of NVP resistance in the 12 women who had NVP resistance detected in breast milk at 4 weeks post-partum (). Eleven of those women had a 10-week sample available for testing; one sample failed to amplify, leaving 10 follow-up samples with genotyping results. NVP resistance mutations were still detected in four of the 10 women by 10 weeks post-partum. The six women who did not have NVP resistance in breast milk at 10 weeks had only one NVP resistance mutation detected at 4 weeks. In contrast, among the four women who still had NVP resistance mutations detected at 10 weeks, three had multiple NVP resistance mutations detected at the earlier time point. Among the four women who had NVP resistance detected in breast milk at 10 weeks, the number of mutations detected at 10 weeks was lower than the number of mutations detected at 4 weeks; in three of the four cases, only K103N was detected in the 10-week sample.
We analyzed the association of breast milk resistance at 4 weeks with clinical and laboratory factors (). NVP resistance was detected in plasma more frequently among women who had NVP resistance detected in breast milk, but the association was not statistically significant (p=0.06, ). We found no significant association of breast milk resistance with median maternal pre-NVP viral load, maternal pre-NVP CD4 cell count, median breast milk viral load at 4 weeks, median pre-NVP white blood cell count, or breast milk sodium >10 mmol/Lat 1 or 4 weeks; none of the 30 women had clinical mastitis. Detection of NVP resistance in breast milk was not associated with HIV subtype.
Factors associated with detection of NVP resistance in breast milk 4 weeks after sdNVP administration (Uganda, 2003–2004).
We also measured NVP concentrations in plasma and breast milk. At 1 week post-partum, NVP was detected (>10 ng/ml) in all 29 evaluable plasma samples and in 28 (96.6%) of 29 evaluable breast milk samples. At 2 weeks post-partum, NVP was detected at >10 ng/ml in 21 (72.4%) of 29 plasma samples and in 16 (69.6%) of 23 breast milk samples. At 2 weeks, detection of NVP in breast milk at >10 ng/ml was associated with detection of NVP in plasma at >10 ng/ml (p<0.001). The median NVP concentrations in plasma and breast milk were similar (at 1 week, for plasma: 168 ng/ml (range 16–744), for breast milk: 117 ng/ml (range <10–576), p=0.16; at 2 weeks, for plasma: 16 ng/ml (range, <10–83), for breast milk: 15 ng/ml (range <10–64), p=0.58). The concentration of NVP was <10 ng/ml in plasma and breast milk from all 30 women by 4 weeks post-partum. There was no association of NVP resistance in breast milk at 4 weeks with either detection of NVP >10 ng/ml in breast milk or plasma, or the level of NVP in either breast milk or plasma, at either study visit (1 or 2 weeks, ).