Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Research Letters AIDS 2006, 20:1891–1901 Antiretroviral activity of didanosine in patients with different clusters of reverse transcriptase mutations Jose Luis Blanco, Alejandra Biglia, Elisa De Lazzari, Josep Mallolas, Esteban Martinez, Toma ´s Pumarola, Marı ´a Larrousse, Ana Milinkovic, Agathe Leo ´n, Mon- tserrat Lonca, Montserrat Laguno and Josep Marı ´a Gatell Patterns of mutations associated with didanosine (ddI) resistance are still a controversial issue. The correlation between different clusters of reverse transcriptase mutations with the short-term viro- logical activity of ddI when added to a failing regimen was examined in 40 patients. The median fall in plasma viral load at week 4 was 0.67 log 10 copies/ml. There was good correlation between the median fall in plasma HIV RNA levels and the number of nucleoside-associated (P ¼ 0.0152) or thymidine-associated (P ¼ 0.0142) mutations. In conclusion, ddI retained substantial antiretroviral activity when the number of nucleoside-associated or thymidine-associated mutations was less than four. Nucleoside analogue reverse-transcriptase inhibitors (NRTI) remain an important component of HAART and have represented the backbone of virtually all antiretroviral regimens. Despite the availability of seven NRTI and tenofovir (a nucleotide analogue), the problem of cross-resistance among this class of drugs is a matter of concern and occurs to a greater extent than initially thought [1]. The recent introduction of didanosine (ddI) in a more convenient and better-tolerated enteric-coated formu- lation makes it an attractive NRTI to be used as part of HAART. Despite 74V being the most common mutation selected by ddI both in vitro and, less frequently, in vivo , other reverse transcriptase (RT) mutations such as the 184V or the 65R can also be selected by ddI when it is given as monotherapy or in sequential NRTI therapy [2]. However, it is unusual to find these mutations in patients failing a ddI-containing HAART, and ddI resistance has been associated with different patterns of mutations not completely defined. Recently the JAGUAR study [3] has shown that ddI retains some antiviral activity against isolates expressing up to three thymidine-associated mutations (TAM) at baseline. Intensification strategies can offer different advantages: first, this strategy may lead to an improvement of the virological control in patients with a low level of virological failure [2,4]; second, it allows the determi- nation of intrinsic antiviral activity of a given drug; and, finally, it is possible to evaluate the patterns of genotypic resistance with a low risk of accumulating new resistance mutations [5]. This short-term intensification study with ddI therapy investigated the relation between the virological response and the presence of different clusters of RT mutations (diNAM study). This was a prospective, open-label single-arm study including 40 consecutive patients under stable antire- troviral therapy not containing ddI or tenofovir and with virological failure (HIV viral load 1000 – 200 000 copies/ ml). After drawing a blood sample for genotypic and virtual phenotype resistance test, ddI (400 or 250 mg daily according to total body weight) was added as a single drug in addition to the baseline antiretroviral regimen for 4 weeks. Clinical, immunological and virological evalu- ation was performed at baseline and at weeks 2 and 4. At week 4, the antiretroviral regimen was optimized according to the results of the resistance tests. Median baseline viral load and CD4 lymphocytes were 4560 copies/ml [interquartile range (IQR) 2810 – 14900] and 405 cells/ml (IQR, 290–505), respectively. At weeks 2 and 4, the median HIV RNA reduction was 0.73 and 0.67 log 10 copies/ml, respectively. At week 4, the proportion of patients with a fall in viral load > 0.5 log 10 copies/ml or > 1 log 10 copies/ml or with a CD4 cell count < 200 cells/ml was 63%, 30% and 24%, respectively. Median rise in CD4 cell count at week 4 was 53 cells/ml. At baseline, the median number of RT mutations, TAM (41L, 67N, 70R, 210W, 215Y/ F, 219E/Q) or nucleoside-associated mutations (NAM: TAM plus 44D and 118I) were four, three and three, respectively. Relations between number of mutations in each cluster (RT/NAM/TAM) and median fall in viral load were 1.37, 0.57, 0.35 and 0.50 log 10 copies/ml for 0–2, 3–4, 5 and 6–7 RT mutations, respectively; 1.2, 0.9, 0.6, 0.4 and 0.1 log 10 copies/ml for 0–1, 2, 3, 4 and 5–6 NAM, respectively; and 1.2, 0.9, 0.7, 0.3 and 0.1 log 10 copies/ml for 0–1, 2, 3, 4 and 5 TAM, respectively. The RT mutations present in > 5% of the patients that showed a significant virological response (P  0.20) comparing with those patients with same codon as wild type were 41L, 118I, 210W and 215Y. Contrasting with the recent analysis of representative mutations in JAGUAR study [6], we did not find significantly better virological response in relation to 70R (a fall of 0.6 log 10 copies/ml in both variant and wild- type codon; P ¼ 0,590) and M184V (a fall of 0.5 versus ISSN 0269-9370 Q 2006 Lippincott Williams & Wilkins 1891