Important new clinical data: potential early virologic failure associated with the combination antiretroviral regimen of tenofovir disoproxil fumarate, didanosine, and either efavirenz or nevirapine in HIV treatment-naive patients with high baseline viral loads

Bristol-Myers Squibb (BMS) issued a “Dear Healthcare Provider” letter in the US, the text of which is included below.

Re: Important new clinical data: potential early virologic failure associated with the combination antiretroviral regimen of tenofovir disoproxil fumarate, didanosine, and either efavirenz or nevirapine in HIV treatment-naïve patients with high baseline viral loads

Dear Health Care Provider,

Bristol-Myers Squibb (BMS) Company is writing to advise you of important new clinical data regarding coadministration of Viread (tenofovir disoproxil fumarate [TDF]), Videx EC (didanosine delayed-release capsules enteric-coated beadlets [ddI/EC]), and either Sustiva (efavirenz [EFV]) or Viramune (nevirapine [NVP]). Data for EFV+TDF+ddI/EC are derived from an open-label randomised study (virologic failure in 6/14 patients) and a retrospective database analysis (virologic failure in 5/10 patients), while data for NVP+TDF+ddI/EC are derived from a retrospective database analysis (virologic failure in 2/4 patients).

Results from two recently conducted, investigator-sponsored trials by Podzamczer et al [1] and JM Gatell (written communication, July 2004) have demonstrated a potential for early virologic failure associated with this antiretroviral regimen in treatment-naïve HIV patients with high baseline viral loads. The mechanism of early virologic failure in these patients is unclear.

Early virologic failure appears to be limited to the specific combination of TDF + ddI EC + either EFV or NVP as there are data from registrational trials supporting the efficacy of EFV and TDF-based regimens as well as EFV and ddI EC-based regimens in treatment-naïve HIV patients. [2, 3, 4]

Additionally, a recent post hoc analysis performed in treatment-experienced HIV patients with high baseline viral loads receiving a boosted protease inhibitor (PI) with two nucleoside reverse transcriptase inhibitors (NRTIs) demonstrated lower virologic failure rates in subjects receiving ddI/EC and TDF than those receiving another nucleoside analogue in combination with TDF, though significance testing could not be performed due to a small number of patients (n=55).

Based on this information:

  • Doctors should use caution when coadministering TDF, ddI/EC, and either EFV or NVP in treatment-naïve HIV patients with high baseline viral loads.
  • Further investigations are ongoing to better understand the clinical implications of these results.

For further details on these studies, please refer to the following pages for study summaries.

Studies demonstrating early virologic failure in treatment-naïve HIV patients with high baseline viral loads

  • The ININ Study [1] (Podzamczer et al): An open-label, randomised, multicenter, pilot study with a planned enrollment of 50 treatment-naïve HIV patients designed to assess efficacy and safety of TDF 300 mg once daily + ddI/EC 250 mg once daily (<60 kg: 200 mg once daily) + EFV 600 mg once daily compared with TDF 300 mg once daily + ddI EC 250 mg once daily (<60 kg: 200 mg once daily) + EFV 600 mg once daily + Kaletra® (lopinavir/ritonavir [LPV/r]) 400/100 mg twice daily. Of the 36 enrolled patients, 26 were available for follow-up at 3 months. Six of 14 patients (42.8%) in the TDF+ddI/EC+EFV arm experienced protocol-defined virologic failure*, versus 0 of 12 patients in the TDF+ddI/EC+EFV+LPV/r arm. Baseline viral load >100,000 copies/mL and advanced stage of disease (low CD4+ cell counts [<200 cells/mm3] plus CDC stage C or B3) were seen in all six patients with virologic failure but in none of the eight patients without virologic failure. Resistance patterns that included G190E/S (n=3), L74V/I (n=4), and K65R (n=2) mutations were observed at failure.
  • JM Gatell et al (written communication, July 2004): A retrospective database analysis of 5,000 treatment-naïve HIV patients in whom therapy was initiated between October 2002 – March 2004 was performed. Fourteen patients were identified as having received a regimen of ddI EC 250 mg once daily and TDF 300 mg once daily, plus either EFV 600 mg once daily (n=10) or NVP 400 mg once daily (n=4). After 12 weeks of therapy, 5/14 patients (36%) experienced suboptimal (plasma viral load drop <2 log10 copies/mL) response rates. Two additional patients (total 7/14, 50%) who were treatment-responders at Week 12 reached protocol-defined virologic failure at Week 24.†The seven cases of virologic failure consisted of 2/4 patients receiving NVP- and 5/10 patients receiving EFV-containing regimens. At baseline, virologic failure patients had a median log10 viral load of 5.8 (range, 4.7-6.0) copies/mL and a median CD4+ cell count of 126 (range, 24-281) cells/mm3. Four of the virologic failure patients exhibited the K65R and L74V mutations and all 7 exhibited one or more of the following mutations: L100I, K103N/R/T, Y181C, and G190E/Q/S.

Studies of treatment-naïve HIV patients with combination antiretroviral regimens containing EFV and TDF or ddI/EC

  • Gilead Study 903 [2]: A 144-week, Phase III, multicenter, randomised, double-blind, active controlled trial in treatment-naïve HIV patients designed to evaluate the efficacy and safety of TDF compared to Zerit (stavudine [d4T]) capsules, each in combination with 3TC + EFV.At 48 weeks, similar efficacy was observed between the two treatment groups: EFV + 3TC + TDF (n=299), HIV RNA <400 copies/mL = 79% and <50 copies/mL = 76%; versus d4T + 3TC + EFV (n=301), HIV RNA <400 copies/mL = 82% and <50 copies/mL = 79%, ITT analysis. At 48 weeks, Study 903 showed comparable virologic efficacy (HIV RNA <400 copies/mL) in patients with baseline viral loads above and below 100,000 copies/mL (n=600; >100,000 copies/mL = 86% in the TDF arm and 85% in the d4T arm; <100,000 copies/mL = 87% in the TDF arm and 89% in the d4T arm). [5] These trends continued through 144 weeks.[6]
  • Study 301A [3,4]: A 48-week, double-blind, active-controlled multicenter study compared Emtriva (emtricitabine [FTC]) 200 mg once daily administered in combination with ddI EC 400 mg once daily (patients weighing <60 kg received 250 mg) and EFV 600 mg once daily versus d4T + ddI EC + EFV in 571 treatment-naïve patients. Thirty-eight percent of patients had baseline viral loads >100,000 copies/mL. In the FTC + ddI EC + EFV arm (n=286), 81% of patients had HIV RNA <400 copies/mL and 78% had <50 copies/mL at Week 48. In the d4T + ddI EC + EFV arm (n=285), 68% of patients had HIV RNA <400 copies/mL and 59% had <50 copies/mL at Week 48.

Study of treatment-experienced HIV patients with a combination antiretroviral regimen containing a RTV-boosted PI, TDF and ddI/EC or another NRTI

  • BMS Study AI424-0457: A Phase III, open-label trial in which 358 treatment-experienced patients with multiple virologic failures were randomised to one of three boosted PIs (shown below) each in combination with TDF and one other NRTI (approximately half on ddI EC):
    • Reyataz (atazanavir sulfate [ATV]) 300 mg + RTV 100 mg once daily or
    • LPV/r 400/100 mg twice daily or
    • ATV 400mg + saquinavir (SQV) 1200 mg once daily.

The ATV + SQV arm had statistically inferior results to those in the ATV + RTV and LPV/RTV arms, and will not be discussed. Interaction studies of ddI EC + TDF demonstrated increased ddI exposure when ddI EC 400 mg was administered one to two hours before TDF 300 mg and a light meal. Consequently, a protocol amendment specified ddI EC dose reduction to 250 mg (adults weighing 360 kg with creatinine clearance 360 mL/min) or 200 mg (adults weighing <60 kg with creatinine clearance 360 mL/min) once daily. By Week 24, approximately 2/3 of ddI EC subjects had reduced dosage to 250 mg.

Forty-eight week results were stratified according to baseline viral load and assessed for differences between ddI- and non-ddI-containing regimens. A post hoc analysis using these data was performed on the subset of treatment-experienced patients with a baseline viral load 3100,000 copies/mL.‡

The two combined arms of ATV + RTV and LPV/RTV demonstrated a 33% (8/24) virologic failure rate§ through Week 48 in the ddI-treated group compared to 52% (16/31) in the non-ddI-treated group.

Please refer to the enclosed full prescribing information for Videx® EC (didanosine) Delayed Release Capsules Enteric Coated Beadlets and Sustiva® (efavirenz) Capsules and Tablets.

BMS is committed to providing you with current product information for the management of your patients with HIV infection. If you have any questions about this new information or require additional medical information, please contact the Virology Medical Services Department at Bristol-Myers Squibb Company at 1-800-426-7644 (select Option 3).


Sally L. Hodder, MD

Vice President, Virology Medical Affairs, BMS

* Virologic failure in the ININ Study was defined as (1) reduction in viral load of less than 2 log10 copies/mL at 3 months, (2) more than 1 log10 copies/mL rebound from the nadir, or (3) viral load detectable at 6 months or later.

† Virologic failure in the Gatell et al study was defined as (1) plasma viral load drop of less than 2 log10 copies/mL at 3 months or (2) viral load rebound less than 200 copies/mL for 2 consecutive measurements separated by at least one week, after an initial drop below 200 copies/mL.

‡ Median baseline HIV RNA: ATV/RTV arm=4.44 log10 copies/mL, LPV/RTV arm=4.47 log10 copies/mL.

§ Virologic failure rates measured by TLOVR (time to loss of virologic response) analysis.


  1. Podzamczer D, Ferrer E, Gatell JM, et al. Early virologic failure with a combination of tenofovir, didanosine and efavirenz. Antiviral Therapy 2004 9:S172, Poster 156,13th International HIV Drug Resistance Workshop; June 2004; Tenerife, Spain.
  2. Viread (tenofovir disoproxil fumarate) Prescribing Information. Gilead Sciences, Inc., June 2004.
  3. EmtrivaTM (emtricitabine) Prescribing Information. Gilead Sciences, Inc., July 2003.
  4. Saag M, Cahn P, Raffi F, et al. Efficacy and safety of emtricitabine vs stavudine in combination therapy in antiretroviral-naive patients: a randomised trial. JAMA. 2004; 292:180-189.
  5. Staszewski S, Gallant J, Pozniak A, et al. Efficacy and safety of tenofovir disoproxil fumarate (TDF) versus stavudine (d4T) when used in combination with lamivudine (3TC) and efavirenz (EFV) in HIV-1 infected patients naïve to antiretroviral therapy (ART): 48-week interim results. XIV International AIDS Conference; July 7-12, 2002; Barcelona, Spain. Oral Presentation LbOr17.
  6. Gallant JE, Staszewski S, Pozniak A, et al. Long-term efficacy and safety of tenofovir DF (TDF): A 144 week comparison versus stavudine (d4T) in antiretroviral-naïve patients. XV International AIDS Conference; July 11-16, 2004; Bangkok, Thailand. Poster 4538.
  7. Data on file, Bristol-Myers Squibb Company, Princeton, New Jersey.


This letter has so far been sent to US but not EU physicians, although both BMS and Gilead are discussing the data with the European regulatory agency and a response will not follow for at least another month.

There were also two studies at ICAAC addressing the same issues that are not referenced in the letter. See ICAAC reports from these studies later in this issue of HTB. Several other studies at the Glasgow conference presented further supportive data.

It is important to separate the virological failure interaction (Podzamczer/CORRS/Gatell), mechanism unknown, from the CD4 data (Negredo) which is likely to be related to a pharmacokinetic interaction, with the discordant CD4 response dependent on ddI dose. As regards low CD4 responses, there is interest with tenofovir inhibiting PNP (since PNP deficiency syndromes in childhood causes SCID-like immunodeficiency) – but low CD4s were not seen in the Gilead 903 study suggesting that this is a ‘high dose ddI plus tenofovir’ problem and not a ‘tenofovir problem’.

Interestingly in the ICAAC poster from Barrios and colleagues (see ICAAC coverage below) patients with a CD4 decline, did not return to baseline even after reducing dose of ddI.

Until these results are more clearly understood, patients starting or switching treatment should clearly avoid the nucleoside combination of tenofovir and ddI.

Patients already successfully suppressed, should still probably switch one or other of these nucleosides, especially if this is a first line combination and convenient alternatives are available. This caution is for risk of cross class resistance should the combination fail in the future and also potential side effects linked to higher ddI exposure, including pancreatitis and lactic acidosis.

Although from a virological perspective this advice is most important in patients whose baseline viral load was >100,000 copies/mL and/or who are less than 100% adherent, the concern for toxicity will apply to all patients.

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