Mutations in the protease gene associated with virological failure to lopinavir-containing regimens in clinical samples

Purpose of the study

Lopinavir/ritonavir (LPV/r) is increasingly being introduced in resource-limited settings as second-line therapy. Data on selection of resistance mutations (RAMs) in treatment-experienced populations failing LPV/r could provide insight into future ramifications of this strategy. We investigated the RAMs in patients with prior or current LPV/r failure.

Methods

We identified 195 samples from multi-treated patients submitted for routine resistance testing. Seventy-one (36%) never received LPV/r, 75 (38.5%) had previously failed LPV/r, and 49 (25%) were currently on LPV/r. Medians, interquartile ranges or percentages, Kruskal-Wallis, χ2 or Fisher test were used whenever appropriate.

Summary of results

Median CD4 282 cells/μl (IQR:252), HIV-1 VL 4.17 log10 (IQR 1.35), and duration of HIV infection 13 years. Number of NRTI, NNRTI and PI RAMs: 5 (IQR:4), 1 (IQR:1), and 6 (IQR:4), respectively. Patients exposed to LPV/r had received more HAART regimens (p < 0.001), PIs (p < 0.001), and NRTIs (p < 0.001). RAMs significantly associated with prior or current LPV/r exposure were: L10I/F (p = 0.02, p = 0.001), K20R (p = 0.023), L24I (p = 0.01), L33F (p < 0.001), M36I (p = 0.029), M46I/L (p = 0.023, p = 0.003), I47V (p = 0.022), G48V (p = 0.08), F53L (p = 0.017), I54V (p < 0.001), A71V (p < 0.001), G73S (p = 0.008), V82A (p < 0.001), I84V (p < 0.001), and L90M (p = 0.037). Two NRTI mutations were also associated with LPV/r failure: E44D (p = 0.002) and V118I (p = 0.048). L76V was found in only one sample. RAMs associated only with current (and not previous) LPV/r failure were: L33F, M36I, M46I, I47V, G48V, A71V, G73S, I84V, L90M. These RAMs are all included in the IAS-USA LPV/r list, except M36I and G48V, found to be associated with prior saquinavir use (p = 0.02 and p < 0.001, respectively). Mutations included in the IAS-USA list for LPV but not found to be associated with LPV/r previous or current failure: L10R/V, K20M, V32I, I47A, I50V, I54L/A/M/T/S, L63P, A71T, I76V, and V82F/T/S.

Conclusion

In multi-treated patients, L33F, M36I, M46I, I47V, G48V, A71V, G73S, I84V, L90M, but not L10R/V, K20M, V32I, I47A, I50V, I54L/A/M/T/S, L63P, A71T, I76V, and V82F/T/S may be over-represented in LPV/r failure. Enrichment of these mutations should be expected in populations receiving widespread salvage with LPV/r. This may assist in considering future options, together with previous use of other PI and subtype prevalence.