To the Editor:
Here we would like to report on an intriguing observation concerning 3 patients with phenotypic and genotypic lamivudine-resistant hepatitis B virus (HBV) infection who suffered from a viral relapse caused by the replacement of tenofovir disoproxil fumarate by adefovir dipivoxil (2 male HBeAg-positive patients, one aged 20 years and one aged 60 years; and one 48-year-old male HBeAg-negative patient, none of them with HIV coinfection). After stopping lamivudine treatment, all 3 patients were directly switched to tenofovir, which is used in the treatment of HIV infection and which also has substantial activity against HBV1 at a daily dose of 300 mg. As shown in Fig. 1, a reduction of HBV DNA to undetectable levels (<2.6 log10 copies/mL, Roche Diagnostic Systems, Pleasanton, CA) could be achieved in 2 of the 3 patients at the end of the observation period of 6 to 7 months. Because adefovir was approved for the treatment of lamivudine-resistant hepatitis B at this time in Germany, we decided to replace tenofovir with adefovir at a daily dose of 10 mg. Unexpectedly, all 3 patients virologically relapsed within the first 4 weeks, and in 2 patients HBV DNA levels rose as high as they were at the beginning of the tenofovir treatment. In the following months, a moderate decrease in HBV DNA could be observed in these 2 patients (Fig. 1), but one of them also showed a marked increase of alanine aminotransferase (ALT) levels within the first 3 months without any further evidence of liver decompensation. Before starting treatment with tenofovir, all patients had proof of lamivudine resistance–associated mutations (rtM204V, affigene HBV DE/3CT, Sangtec Molecular Diagnostics, Bromma, Sweden) within the HBV polymerase gene which became undetectable during the treatment with adefovir. Moreover, we could exclude the previously described genotypic resistance to adefovir (rtN263D/T or rtA181V/T) by direct sequence analysis of the entire HBV polymerase gene (Shelly Xiong, Gilead Sciences, personal communication, December 2004).
How can we explain the failure of adefovir to retain the tenofovir response? In vitro studies suggest that tenofovir and adefovir are equipotent on a molar basis.2, 3 Thus, 300 mg tenofovir is likely to be more active in vivo than 10 mg adefovir, and therefore dosages above 10 mg could have prevented these relapses. Additionally, variable antiviral effects of adefovir definitively must be taken into consideration, as recently demonstrated.1 Finally, differences in the intracellular metabolism of the two drugs may be involved, as it was shown that tenofovir achieved a higher intracellular diphosphate concentration than adefovir in HepG2 cells.4, 5 However, one should be aware that acyclic nucleotide analogs may be able to express several other putative mechanisms of action, including that of an immunoregulator.6–10
From these preliminary observations, a change from one nucleotide analog to another may be risky. It still remains speculative whether overlapping regimens are indicated under these circumstances, especially since the antiviral properties of these drugs are not clearly defined.