Characterization on the dynamics of disappearance of protease resistance mutations during treatment interruption

Poster number: 18

MM Santoro (1), C Gori (2), V Svicher (1), M Santoro (1), F Forbici (2), N Esposito (1), M Zaccarelli (2), A Cenci (1), A Antinori (2), CF Perno (1,2), F Ceccherini-Silberstein (1)

  1. University of Rome Tor Vergata, Rome
  2. INMI L Spallanzani, Rome, Italy

Background: The dynamics of disappearance of resistant variants during treatment interruption (TI) varies according to the number and type of mutations, and to the presence/absence of other mutations able to increase the replicative capacity of the virus. We investigated the evolution of mutations associated with resistance to protease inhibitors (PI) during TI, and their ability to affect virus replication.

Methods: Analysis on 88 patients having at least two genotypes, one at the last regimen failure containing PI (overall 89 genotypes) and one (or more than one) during TI (overall 113 genotypes), with a delta <1 year from genotype at last HAART-virological failure and start of TI, was performed. TI lasted from 1 to 19 months (median, 4 months). The association between mutations and viremia/CD4 cell-count was assessed by median test. Chi-square test was used to detect whether specific mutations (present at TI) affect the dynamics of the disappearance of other mutations. The time of disappearance (mean life) t of major PI-mutations was estimated by fitting an exponential decay for their number (N = N0 exp(-T/t)) in 3 independent groups of patients, with high (>3), medium (2-3) and low (1) major mutations number (N0) at the last HAART-virological failure (assuming N0 as constant at the start of TI (T=0), in 61 patients having a genotype at last HAART-virological failure with delta time <90 days before T=0). The time of disappearance t of a specific mutation was also estimated by fitting an exponential decay (F(t)=F0 exp(-T/t)) for the mutation frequency alone or in the presence of other mutations.

Results: Complete disappearance of major PI-resistance mutations and of recently associated-resistance mutations (K20T, K43T, Q58E, T74S, I85V, Q92K, C95F) occurred between 6 and 12 months of TI; this strongly supports their role in drug-resistance. By contrast, minor mutations M36I, L63P, V77I remained stable with high frequency even after TI (>75%), suggesting a limited interference with viral fitness.

The disappearance of M46I and L90M during TI was associated with higher viral load increase and with greater decrease of CD4 cell-count compared with viral load from patients that maintained such mutations.

The mean-life of L90M (134+/-6 days) was prolonged by the specific presence at baseline of M36I and K20R mutations (from 107+/-6 to 330+/-40 days, and from 127+/-4 to 240+/-30 days without/with M36I or K20R, respectively). Regarding the disappearance of M46I (t = 122+/-3 days), the absence at baseline of L10I had determined a faster disappearance (from 122+/-5 to 29+/-4 days with/without L10I), while also after 240 days of TI, in the presence of M36I the M46I mutation did not disappear.

Conclusions: Selected minor mutations play a critical role in the appearance and stabilization of M46I and L90M at virological failure. Overall, the characterization of mutations more detrimental for viral fitness can be used in designing clinical strategies aimed to reduce the virus-related damage in highly drug-experienced patients.