709 
Type-1 Thymidine-associated Mutations but Not K65R Mutation Play a Role in Determining Virologic Failure to Combined Rescue Therapy with Tenofovir and Stavudine
Andrea Antinori*1, M Trotta1, P Nasta2, T Bini3, S Bonora4, A Castagna5, T Quirino6, S Landonio3, S Merli3, V Tozzi1, M Zaccarelli1, G Di Perri4, M Andreoni7, C Perno7, and G Carosi2
1INMI, IRCCS L Spallanzani, Rome, Italy; 2Univ degli Studi, Brescia, Italy; 3Hosp Luigi Sacco, Milan, Italy; 4Univ degli Studi, Torino, Italy; 5Hosp San Raffaele, Univ Vita e Salute, Milan, Italy; 6General Hosp, Busto Arsizio, Italy; and 7Univ of Rome 'Tor Vergata', Italy
Background: Tenofovir (TDF) and thymidine analogues seem to
determine different pathways of viral evolution under selective pressure on HIV
reverse transcriptase.
Methods: Multicentric
retrospective analysis on patients
starting TDF/stavudine (d4T) after HAART failure. Linear regression analyses
were used to determine predictors of HIV RNA change at 6 months of therapy;
risk of virologic failure (time to first HIV RNA > 500 copies/mL occurred
after 3 months of therapy) was calculated using KM estimate, and its
determinants by multivariate Cox model.
Results: Of 172
patients, all but 2 had been previously exposed to thymidine analogues, zidovudine
(AZT) in 88% and stavudine (d4T) in 77%. The mean viro-immunological values
were 4.26 log10/mL and 229 cell/mm3 for HIV RNA and CD4
cell count, respectively. At baseline, a genotype was available in 136 patients;
the most frequent mutations were: M184V
(76%), T215Y/F (42%), M41L (33%), D67N (29%), and L210W (22%); 33% had at least
3 thymidine analog mutations (TAM).
Any single type-1 TAM mutation
(M41L: β = +0.969, p = 0.001; L210W: β = +0.967,
p = 0.003; T215Y: β = +1.250, p < 0.0001) had a
negative effect on the change in HIV RNA at 6 months, whereas among type-2 TAM mutations, only D67N showed a trend for a negative
effect (β = +0.582, p = 0.051), both K70R (β = –0.027,
p = 0.934), and K219Q (β = –0.938, p = 0.083) had no
effect. Presence of M184V mutation was also related with a greater reduction in
HIV RNA (β = –0.759, p = 0.019). The KM estimated risk of viral
failure at 6-months was 22%. Multivariate Cox model confirmed the detrimental
effect of type-1 TAM after
adjustment for several variables: for each more type-1 TAM
mutation we found a greater risk of viral failure (AHR
= 1.65; 95% CI 1.19 to 2.29; p = 0.003),
and the presence of the entire pattern had a 2-times risk of viral failure (AHR = 2.53; 95% CI 1.01 to 6.33; p = 0.05). Conversely, M184V was
associated with a protective effect (AHR
= 0.36%CI 0.15 to 0.87; p = 0.02). In
13 GRT at viral failrue, no K65R
mutation was detected, whereas a trend for an increasing prevalence of
d4T-associated mutations was found (D67N from 23% to 46%; V118I from 15% to
31%; L210W from 31% to 67%; K219Q from 7.7% to 15%).
Conclusions: In patients
previously exposed to nucleoside reverse transcriptase inhibitors (NRTI), GRT at baseline may predict virological success to
rescue therapy combining d4T with TDF:
type-1 TAM negatively affect virological outcome, while M184V seems to
have a favorable effect. Accumulation of TAM, but not K65R selection, is the
most relevant GRT pattern at
virological failure to TDF/d4T combination.
Keywords: HIV drug resistance; TAMs; K65R
