Thymidine analogues mutations (TAMS) and other NA-related resistance mutations

Several presentations addressed the relationship between ZDV-associated mutations and d4T resistance, the issue of thymidine analogue mutations (TAMS) and resistance patterns associated with multi-nucleoside reduced susceptibility.

The mechanism of d4T resistance is poorly understood. It generally develops more slowly than for other nucleoside analogues, and reduced virological sensitivity with most patients is not associated with a clear pattern of genotypic mutations. These studies emphasised the complexity of phenotypic and genotypic interactions and that closer developed research is still needed to understand implications for clinical practice.

ZDV-mutations in d4T exposed patients

In a study presented by Anton Pozniak, from the Chelsea and Westminster Hospital, London, a clinical and genotypic database of 640 patients was used to find patients with ZDV-associated mutations, who had not previously used ZDV [1]. 227/640 patients were identified as using d4T with no prior ZDV-use, and genotyping from these samples found 35 people with ZDV-like mutations. Included in this group were two people who were no longer using d4T in their regimens and two people who had discontinued treatment but who had also previously used d4T.

Genotype results showed 15/35 to have single ZDV-associated mutations. 7 samples had 2 mutations and 6, 6 and 1 patients had 3,4 and 5 mutations respectively. The most common substitutions were T215F/Y (17/35), M41L (14/35), D67N (13/35) and K70R (11/35). However, only 10 of the 26 patients with changes at positions 70 or 215, showed significant phenotypic resistance to ZDV that these changes would be expected to predict.

Table 1. Phenotypic resistance: sensitivity to ZDV

Sensitivity n %
< 4-fold 15 42.8
4-10-fold 7 20%
> 10-fold 10 28.5
No result 3 8.5%

The study did not present probability analysis for these results nor details of baseline resistance, which are currently being analysed. Given that conservative estimates of 5% rate of infection with ZDV resistant virus have regularly been reported for the last few years in urban US and European populations, this could complicate the 15% (35/227) detected in this adult population.

The investigators noted that levels of phenotypic resistance to ZDV were reduced when these mutations developed through d4T- rather than ZDV-based treatment. This may point to an additional resistance-associated benefit for using d4T as a first-line nucleoside. It was also noted in a similar study presented by Graeme Moyle from the same clinic, that a large percentage of the overall d4T-exposed population do not develop ZDV-like mutations and that his would suggest ZDV would remain an option for future treatment in these individuals. [2]

Further evidence of TAMs associated with d4T was presented by Rob Shuurman from the University Medical School, Utrecht, in a retrospective analysis of samples from PACTG240 (a paediatric d4T monotherapy study). In this study, pre-existing baseline resistance to ZDV was measured – and would have been unlikely to be a complicating factor due to negligible chance of exposure in this patient group. [3]

32 paired samples (one each from start and end of treatment) from 16 children who used d4T monotherapy for a mean 75 weeks were genotypically analysed. 11/16 children were antiretroviral na•ve, 4 had been exposed to ZDV for <5 weeks and the mother of one child had received ZDV during pregnancy. All other mothers were antiretroviral na•ve through pregnancy. In vitro replication was investigated in HxB2-infected SupT1 cells that were cultured in the presence of increasing concentrations of d4T from 2.5µM (approximate IC50 value) -80µM, and RT sequenced following several passages in the presence of 80µM 4dT.

None of the patients showed ZDV-mutations at baseline. Five patients developed 1-3 mutations associated to ZDV resistance (41L, 70R, 210W. 215Y/F) although no multi-drug resistant mutations were detected. Changes in the IC50 were low in the children without ZDV-associated mutations (mean IC50 – 5.4µM) – although this still indicates a 2-fold change, but correlated significantly with the number of ZDV-associated mutations (mean IC50 – 10.7µM – approximately 4-fold). The in vitro selection experiments confirmed the selection of ZDV mutations (at codons 215, 69 and 75) in some of those samples.

d4T and background nucleosides

Vincent Calvez from Pitie Salpetriere Hospital, Paris looked at patterns of resistance in treatment naive patients treated for over 6 months with d4T/3TC (dual only), or d4T/3TC or d4T/ddI within a triple combination. [4] In the patients analysed who had used only d4T/3TC dual therapy (n=44), only one patient selected for T215Y and one patient selected for MDR-resistant Q151M by week 24-48, both of whom had tested wild-type at baseline. All 44 patients in this group had developed the M184V 3TC-associated mutation over the same period.

None of the patients who had tested wild-type at baseline who used d4T/3TC (n=15) as part of a triple combination developed ZDV/TAM-related mutation whatever the combination (whether PI or NNRTI-based). Once again, the 184V mutation was detected in all patients, together with NNRTI-associated mutations in some patients on NNRTI-based therapy.

TAMs were present in 3 patients using d4T/ddI (n=18) as part of a triple combinations (and not as frequently as indicated in the previous dual-therapy ALBI, STADI, BMS A1460 studies) and the time of exposure to d4T/ddI was reported as being an important parameter in this.

d4T and abacavir in vitro

Further in vitro data was presented by Annick Ruffault, Pontchaillou Hospital, Rennes on phenotypic sensitivity to both abacavir and d4T in plasma RT viral genotypes of 21 ZDV/3TC-experienced patients (ZDV>36 months, 3TC>6months) [5]. The results were separated into Group A: patients with 2 ZDV-associated mutations + M184V (n=5); and Group B: patients with >=3 ZDV-associated mutations + M184V (n=14) or + E44A/V118I (n=2). Phenotypic sensitivity was defined as < 4-fold = sensitive (S), 4-8-fold = intermediate (I) or > 8-fold = resistant (R).

Group A showed 100% as sensitive to d4T and 80% sensitive/20% intermediate to abacavir. Group B showed 50% S, 38% I and 12 R to d4T and 63% S, 25% I and 12% R to abacavir.

The study suggested that phenotypic testing was more useful than genotypic for identifying d4T and abacavir na•ve patients who are unlikely to benefit from these drugs (ie who show >4-fold loss in sensitivity). One caution against positively identifying patients sensitive to d4T and abacavir from this in vitro study, would be the variable definitions of phenotypic sensitivity. It has been suggested elsewhere, for example, that both d4T and abacavir may lose antiviral activity at with lower than 4-fold sensitivity and that the scale of phenotypic sensitivity for individual drugs may require individually defined scales of reference.


Currently, the two major commercial phenotypic assays have different fold-resistance cut-offs for defining drug sensitivity (Virologic = <2.5-fold, Virco = <4-fold), which can lead to confusion in interpreting results between studies and in determining the implications for clinical practice. More importantly, if smaller changes in phenotypic sensitivity are more important for some antiretroviral agents than for others, then it will not be possible to rely on a single fixed rage of sensitivity cut-offs to cover all drugs.

Resistance to d4T generally occurs more slowly than with other nucleoside analogues, but If, when it does develop, a 2- or 3-fold reduction in sensitivity to d4T is sufficient to produce reduced virological activity, this may explain why genotypic mutations are less frequently recorded. Mutated virus may not become the majority population in this situation, or, if insufficient selective pressure is limiting the development of further resistance, this may limit/reduce the risk (except in the case of MDR mutations) of developing cross-resistance to other class compounds.


The evidence is accumulating that these two thymidine analogues share common resistance patterns. Pathways to resistance may, however, differ as do their tolerability and side-effect profiles. Further studies are needed to determine if there may be benefits in using either agent as first line or second line therapy.


  1. A Pozniak, Y Gilleece, M Nelson et al. Zidovudine genotypic and phenotypic resistance arising in patients never exposed to zidovudine. 4th International Workshop on HIV Drug Resistance & Treatment Strategies. 12-16 June 2000, Sitges, Spain. Abstract 56.
  2. G Moyle, B Gazzard. Differing reverse transcriptase mutation patterns in persons experiencing viral rebound on first-line regimens with stavudine/didanosine and stavudine/lamivudine. 4th International Workshop on HIV Drug Resistance & Treatment Strategies. 12-16 June 2000, Sitges, Spain. Abstract 28.
  3. R Schuurman, M Nijhuis, W Keulen et al. Selection of zidovudine resistance mutations confering low-level resistance to stavudine occurs at low frequency in stavudine-treated patients and in vitro during prolonged selection experiments. 4th International Workshop on HIV Drug Resistance & Treatment Strategies. 12-16 June 2000, Sitges, Spain.Abstract 53.
  4. V Calvez, M Mouroux, D Descamps et al. Occurrence of Thymidine-Associated Mutations (TAMs) in Naive Patients Treated more than 6 months by Stavudine/Lamivudine Bi-Therapy Combination and Tri-Therapies including Stavudine/Didanosine of Stavudine/Lamivudine. 4th International Workshop on HIV Drug Resistance & Treatment Strategies. 12-16 June 2000, Sitges, Spain. Abstract 54
  5. A Ruffault, P Jaccard et al. Impact of zidovudine-lamivudine resistance mutations on phenotypic susceptibility of HIV-1 isolates to abacavir and stavusine using a HeLa LacZ CD4 CCR5 cell indicator assay. 4th International Workshop on HIV Drug Resistance & Treatment Strategies. June 2000, Sitges, Spain. Abstract 61.

Links to other websites are current at date of posting but not maintained.