INTRO: Protease inhibitor (PI) resistance mutations have not been reported in semen HIV-1 RNA. Semen and blood protease (PR) and envV3 of patients (pts) on amprenavir (APV) therapy (Rx) were sequenced.

METHODS: 9 previously PI-naïve pts were studied: 7 on APVmonoRx, 2 on APV/ZDV/3TC. 7 had baseline semen plasma>blood plasma (BP) viral load (VL, Nuclisens; 400 c/ml limit); 4 with >106c/ml. Four had semen VL rebound on Rx. PR & V3 were RT PCR'ed from DNase Rx'ed RNA, spearate PCRs pooled, cloned, & sequenced(ABI 377).

RESULTS: The first major PI resistance mutations in semen plasma RNA were seen in 3 APV monoRx pts. Two hadAPV-selected mutations (50V or 10I) and rebound in both Semen & BP RNA. A third had 90M and VL rebound in semen, but not in blood.V3s have been analyzed for 3 pts: those with 50V or 90M in rebounding semen and one pt with 50V on-Rx only in blood without semen VLrebound. Significantly different V3s were in each pts' semen and BP prior to Rx and remained different on Rx in the pt with no semen VLrebound. However, on-Rx semen V3s became more related to blood V3s in both pts with semen VL rebound and PR resistance. Minimal semen V3 diversity was seen, even when baseline semen VL > 106 copies/ml.

CONCLUSIONS: PI resistance mutants can be seen in semenwith rebounding VL on Rx, with or without BP VL rebound. There is some PI selective pressure in semen, although V3 genetic diversity islimited. Semen and blood V3s can become more related on Rx during semen VL rebound. Drug selection may involve trafficking betweenblood and male genital tract.