icon star paper   Articles  
Back grey_arrow_rt.gif
Tenofovir Resistance in Treatment-Naïve: study 903 tenofovir vs d4T
Reported by Jules Levin
  View the: Tenofovir Resistance in Treatment-Naïve: study 903 tenofovir vs d4T Slide Show
The current study is the first analysis presented on the development of tenofovir resistance in treatment-naïve patients. Study patients are receiving tenofovir or d4T therapy in combination with lamivudine (3TC) and efavirenz.
This report is a revision of my previous commentary in a report I emailed on this study, in which I discussed the significance of the emergence of the K65R tenofovir mutation in patients in this study experiencing virologic failure on tenofovir.
This article reports on Michael Miller's (Gilead Sciences) poster presentation at the Glasgow HIV Conference in November 2002 on a resistance substudy of Study 903. Study 903 is an ongoing Phase III clinical trial of tenofovir (Viread) or d4T (stavudine) in combination with 3TC and efavirenz (Sustiva). In this interim analysis, plasma HIV (in the blood) viral load was analyzed from all patients with virologic failure during the first 48 weeks (ITT analysis: intent-to-treat). Genotypic analysis was performed by Virco and phenotypic testing was performed by both Virco (Antivirogram) and Virologic (Phenosense). Virologic failure was defined as patients with >400 copies/ml of HIV RNA at week 48 or at early discontinuation for any reason.
Before taking a look at the Glasgow presentation by Miller, here's a look at previous studies exploring tenofovir resistance. In vitro (in the test tube), the K65R mutation can be selected by tenofovir and results in reduced susceptibility to tenofovir and some other NRTIs. In studies of treatment-experienced patients receiving tenofovir the K65R mutation does not appear to emerge easily or often. In study 907, a study in 552 treatment-experienced patients, it was found that 3% of the study patients developed the K65R mutation. The presence of the K65R mutation or several thymidine analogue mutations (AZT-associated mutations) reduced susceptibility to tenofovir. In study 902, 2% of study patients developed the K65R but continued to have significant viral suppression. A search of PMPA or tenofovir on the NATAP website will yield numerous reports, including its potential use as a topical gel in preventing HIV transmission. Here is a link to read about the tenofovir resistance studies in more detail: Studies 907 and 902- http://www.natap.org/2001/ICAAC/day4.htm Study 902 resistance report http://www.natap.org/2000/7thcroi/46rpt30500.html
There are two nucleoside mutations that are rare but can cause cross-resistance to multiple nucleosides. The T69S double insertion multi-nucleoside resistance mutation is rare but can cause resistance to tenofovir. The Q151M multi-nucleoside mutation is also rare and does not appear to cause cross-resistance to tenofovir.
Resistance to HIV drugs in general is a continuum, it is not black and white. A patient may have some resistance to a drug but still may retain some sensitivity or susceptibility to the drug. This applies to tenofovir as well. Using phenotypic resistance testing, the mean response to tenofovir in study 907 was reported to be -0.72 log when there was no phenotypic resistance at baseline, -0.50 log when there was 1-2 fold phenotypic resistance, -0.42 log when there was 2-3 fold phenotypic resistance, -0.27 log when there was 3-4 fold resistance and 0.08 log when there was >4 fold resistance. In treatment-experienced patients with genotypic resistance to NRTIs, tenofovir has thus been shown to be effective. Briefly, in trying to assess viral load response expected when using tenofovir in treatment-experienced patients, the more AZT-associated mutations that are present the less of a viral load reduction can be expected. In a study of treatment-experienced patients, when 3 or more of these mutations were present including the M41L or L210W (AZT mutations) the viral load reduction after adding tenofovir to a patient's regimen was -0.21 log. However, when three or more AZT-associated mutations were present without M41L or L210W, the viral load reduction was -0.67 log. In small studies of treatment-naïve patients, viral load reduction associated with tenofovir was 1.5 log. Here is link to recent tenofovir resistance report from the 11th Intl HIV Drug Resistance Workshop (Summer 2002): http://www.natap.org/2002/HIVDrugRes/day4.htm
Gilead Study 903 is a placebo-controlled trial comparing tenofovir 245 mg (300 mg pill) daily with d4T 40 mg daily (weight adjusted, immediate release formulation) in 600 treatment-naïve patients. 25% of study participants were female and 36% were non-white. At baseline, before starting study drugs, mean HIV viral load was 81,300 copies/ml and CD4 count was 280. About 40% of patients in this study had more advanced HIV at baseline: 43% of patients had a high viral load (>100,000) copies/ml viral load and 39% had <200 CD4s. Despite this, the patient response to therapy was good: 82% of the patients receiving tenofovir and 81% of the patients receiving d4T had <50 copies/ml at week 48 using the intent-to-treat "missing equals failure" analysis (the more stringent analysis). 87% of patients had <400 copies/ml at week 48 whether they received d4T or tenofovir, also using the intent-to-treat "missing equals failure" analysis. CD4 increases were comparable, 169 cell increase for all study patients at week 48. The study is continuing to follow these patients for two additional years.
In the tenofovir group, 29 individuals (9.7%) were defined as virologic failures and in the d4T group 25 individuals (8.3%) were defined as virologic failures. Mutations associated with efavirenz resistance were found in 16 tenofovir failures and 12 d4T failures. The 3TC M184V/I mutation was found in 12 tenofovir failures and 8 d4T failures. Wild-type virus (no resistance detected), was found in 10 of the 29 tenofovir failures and 12 of the 25 d4T failures.
As mentioned above, the K65R mutation has been found in the test tube to decrease tenofovir susceptibility, and this observation has also been seen in patients in clinical study. None of the patient failures in either treatment group in study 903 was found to have the K65R alone. But, the K65R mutation was seen along with efavirenz resistance in 2 tenofovir failures and 1 d4T failure, and in combination with efavirenz and 3TC resistance in 5 tenofovir failures and 1 d4T failure. Overall, 7 of 29 (24%) tenofovir failures developed the K65R mutation and 2 d4T failures (8%) developed the K65R mutation. In speaking with Gilead researchers they said efavirenz and 3TC resistance mutations developed first followed by the K65R mutation. This suggests to me that the K65R mutation is difficult to develop and perhaps is more likely to develop when higher levels of viral replication are present.
The blood samples, or as often called the patient isolates, of the 7 patients with tenofovir failure and who developed the K65R mutation were evaluated for phenotypic resistance using both the Virco and Virologic phenotypic tests. The results were one of the more interesting aspects of the study as the tests sometimes reported different results highlighting the difficulties in using and interpreting phenotypic resistance test results. Using the Virco Antivirogram test, the cutoff for tenofovir resistance is 3.0 fold resistance. Four of the 7 tenofovir failures with the K65R had greater than 3.0-fold tenofovir resistance (patient #1 had 5.9 fold resistance; patient #3 had 3.9 fold resistance; patient #5 had 3.4 fold resistance; patient #7 had 6.4 fold resistance). Three of these 4 patients had the M184V mutation in addition to K65R. The other 3 patients had <2.0 fold tenofovir resistance; 1 of these patients also had the K65R in combination with M184V. Using the Virco assay, 3 patients (patients 1, 2, and 7) had abacavir resistance (>3.0-fold abacavir resistance), 3 (patients 1, 3, and 5) had ddI resistance (>3.5-fold ddI resistance), and none had d4T (>3.0-fold d4T resistance) or AZT resistance (>4.0-fold AZT resistance). Five patients appeared to have hypersensivity to AZT (<1.0-fold resistance). All 7 tenofovir failures with the K65R mutation started a new regimen and 2 of the 7 tenofovir failures retained tenofovir in their second line therapy regimen and these 2 patients achieved <50 copies/ml.
Using the Virologic phenotype test, the cutoff for tenofovir resistance is 1.4- fold resistance, while 3.0 is the cutoff using the Virco test. In contrast to results from using the Virco test, 6 of the 7 patients had less than 1.4 fold tenofovir resistance.
As noted above, four patients had the K65R mutation in combination with the M184V mutation (patients # 1, 2, 3 and 7). Using the Virologic phenotype assay, 2 patients (patients # 1 and 2) had abacavir resistance (>4.5-fold resistance), and 3 patients (patients # 1, 2 and 7) had ddI resistance (>1.7-fold resistance). No patients had d4T or AZT resistance. Again, several patients appeared to have AZT hypersensitivity. The AZT resistance cutoff using the Virologic assay is 2.5-fold AZT resistance and the d4T cutoff is 1.7-fold d4T resistance.
As you can see the findings from two phenotypic tests have a number of differences between them. Although the results were consistent for AZT and d4T, both showing no evidence of resistance with the K65R mutation, resistance for other NRTIs is at times detected using one assay but not with the other assay. The results from the two tests highlight some of the difficulties in using and interpreting phenotypic resistance testing.
From in vitro phenotypic analysis with the Virologic assay, the study authors reported that patients who did not have 3TC resistance (M184V mutation) retained full susceptibility to abacavir. In contrast, 2 out of 4 patients who had the K65R and M184V had abacavir resistance, 1 patient remained susceptible to abacavir and data was unavailable for 1 patient. So 2 out of three patients who developed the K65R and M184V and for whom data was available had abacavir resistance. Using the Virco assay all 3 patients who had abacavir resistance had M184V and K65R mutations but there was 1 patient with this combination of mutations who remained susceptible to abacavir. Patients who only had the K65R without M184V were fully susceptible to abacavir as measured by both the Virco and Virologic assays. One patient with the K65R and the 184 had 10 fold ddI resistance using the Virco test but was fully sensitive (0.9 fold resistance) to abacavir. You have to be careful using resistance testing. It's very important that a doctor be very knowledgeable about resistance testing when interpreting test results. Even the best resistance experts often differ on the interpretation of test results.
The authors reported follow-up on the 7 patients with the K65R and tenofovir failure. The median follow-up was 50 weeks. All 7 patients were switched to a new regimen and 5/7 achieved <50 copies/ml. One patient had 423 copies/ml and was lost to follow-up and a second patient had 1900 copies/ml and was non-adherent. Interestingly, 2 patients (patients # 1 and 2) with the K65R and M184V retained tenofovir in their new regimen and both achieved <50 copies/ml. Patient 1 was called to be resistant to tenofovir by the Virco assay and sensitive by the Virologic assay. Patient 2 was called to be sensitive to tenofovir by both assays. Patient 1 was at week 32 in follow-up, and patient 2 was at week 36 in follow-up. Longer follow-up of these 2 patients would be interesting. The new regimens were selected by the patient's physicians. Interestingly, 1 patient was placed on two drug therapy: ddI plus indinavir. You have to question this because there is no evidence that 2 drugs are adequate but according to the study results this patient remained <50 copies/ml in follow-up.
View the: Tenofovir Resistance in Treatment-Naïve: study 903 tenofovir vs d4T Slide Show
  icon paper stack View Older Articles   Back to Top   www.natap.org