Pre-existing mutations related to tenofovir in chronic hepatitis B patients with long-term nucleos(t)ide analogue drugs treatment by ultra-deep pyrosequencing

Abstract

Aims: The dynamics of resistance-associated mutations under combination therapy were explored.

Methods: A total of 46 patients were classified into adefovir (n=14) and entecavir (n=32) groups. In the adefovir (ADV) group, six patients receiving combined therapy were DNA-positive after more than 3 years of therapy. Ultra-deep pyrosequencing was used to analyze the dynamics of multi-drugs resistance mutations.

Results: At baseline, all 46 treatment-naïve patients harbored rtA181V/T substitutions (1.2%-4.6%) and rtN236T substitutions (1.6%-6.1%). In the ADV group, eight patients with long-term treatment were consecutively HBV DNA-positive for more than 3 years. During treatment, the rtA181T resistance-associated site appeared with increasing frequency in six of eight patients (NOs. 1-6), and two patients (NOs.4 and 8) carrying the rtA181T resistance mutations increasingly showed high levels of rtN236T. One patient (NO. 8) experienced virological breakthrough. Other known pre-existing mutations showed no dynamic fluctuations, including in rtA194T, rtP177G, rtF249A, and rtD263E. In addition to the common substitutions, some previously unknown amino acid substitutions, such as rtD134N, rtL145M/S, rtF151Y/L, rtR153Q, and rtS223A, should be further studied.

Conclusions: HBV-resistance substitutions conferring to nucleoside analogs are present at baseline. The dynamics of the HBV RT-region quasispecies variation are heterogeneous and complex.

Keywords: hepatitis B virus; multi-drugs therapy; resistance; tenofovir; ultra-deep pyrosequencing.

Figure 1. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV+LdT treatment in patient 1

A. The level of serum HBV DNA declined gradually and was undetectable after 4 years of treatment and 7 years of followed-up. B. The dynamic changes in the RT domains of HBV variants, as derermined by UDPS. At baseline, this patient harbored the rtL180M (2.4%), rtM204I/V (3.4%), rtS202G (3.3%), rtA181V/T (4.7%), and rtN236T (6%) substitutions. The resistance-associated variants show minor fluctuations after the addition of LdT therapy in the first year. A wave of resistance-associated variants was detected at year 3, when the frequency of the rtA181T substitution rose from 1.91% to 54.23%. Meanwhile some certain non-resistance-related mutations of rtL145M (ranging from 76.77% to 7.30%) and rtF151Y (ranging from 72.61% to 7.15%) had decreased significantly in frequency with the addition of LdT. However, the frequencies of the rtS223A and rtN248H mutations declined in the first year with the addition of LdT and then increased significantly in the second year.

Figure 2. The dynamic changes in HBV DNA levels and resistance-associeted variants in the RT domains during ADV+LdT treatment in patient 2

A. The serum HBV DNA level was undetectable after 4 years of treatment. B. Patient 2 added LdT to after treatment with ADV for 1 year. At baseline, this patient harbored the rtA181T (4.7%), rtN236T (5.7%), rtL180M (12.5%), rtS202G (3.2%), and rtM204I/V (7%) substitutions. The frequencies of the resistant variants show minor fluctuations after the addition of LdT one year after beginning treatment. The frequency of the rtA181T substitution increased to 75.41% in the first year with the addition of LdT, then declined to 28.07% in the second year. In contrast, the rtQ215R substitution increased to 10.89% in the second year with multi-drugs. Some certain non-resistance mutations of rtD134N (ranging from 91.97% to 0.21%), rtL145M (ranging from 78.58% to 0.27%) and rtF151Y (ranging from 74.89% to 0.27%) had decreased significantly with the addition of LdT. However, the frequency of the rtS223A mutation increased significantly (ranging from 18.14% to 77.62%) with the addition of LdT.

Figure 3. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV+LdT treatment in patient 3

A. The serum HBV DNA level declined gradually and was undetectable after 5 years of treatment. B. This patient added LdT at the end of the 2 years. At baseline, this patient harbored the rtA181T (1.4%), rtN236T (1.8%), rtS202G (1%), and rtM204I/V (1%) substitutions, and other resistance mutations were present at low levels (<1%). The frequencies of the rtA181T and rtN236T mutations increased during ADV monotherapy and declined when LdT was added, but the fluctuations were minor. Some non-resistance-associated mutations of rtD134N (ranging from 20.33% to 74.63%), rtL145M (ranging from 2.83% to 78.82%), rtF151Y (ranging from 2.92% to 75.51%) and rtS223A (ranging from 5.77% to 18.44%) increased significantly with ADV monotherapy, then declined with the addition of LdT.

Figure 4. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV+LdT treatment in patient 4

A. The serum HBV DNA level was undetectable after 6 years of treatment. B. This patient added LdT at the end of the second year. At baseline, patient 4 harbored the rtA181T (1.2%), rtN236T (1.4%), and rtV214A (1.2%) substitutions, and other resistance mutations were present at low levels (<1%). The frequencies of the rtA181V and rtN236T mutations in the HBV RT region were significantly elevated (up to 84.35% and 67.67%, respectively) after two years of ADV monotherapy and did not decreased after the addition of LdT. The frequencies of the rtL180M and rtM204V mutations fluctuated slightly in combination with LdT.

Figure 5. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV+LAM treatment in patient 5

A. The serum DNA level declined gradually and was undetectable after 5 years of treatment with the addition of LdT after 1 years and 9 months. B. This patient harbored the rtA181T (1.2%), rtN236T (1.8%), rtS202G (1.1%), and rtM204I/V (1.1%) substitutions, and other resistance mutations were present at low levels (<1%) at baseline. The frequency of the rtA181V mutation significantly fluctuated (ranging from 57.34% to 1.14%) after combination with LdT. This phenomenon usually appeared in non-resistance mutations of rtS223A (ranging from 79.67% to 4.41%) and rtN248H (ranging from 72.38% to 5.03%).

Figure 6. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV+LAM treatment in patient 6

A. The serum DNA level declined to 10⁴ log10 IU/ml with 2 years of ADV monotherapy. Then, after combination with LAM at the end of the 2 years, the serum DNA level increased to 10⁶ IU/ml. Thus, in this patient treated with ETV and ADV, the HBV DNA level increased to 10⁷ IU/ml after 1 year. B. At baseline, this patient harbored the rtA181T (8.7%), rtN236T (1.7%), rtL180M(10.5%), rtM204I/V (7.5%), and rtV173A/M (1.6%) substitutions, and other resistance mutations were present at low levels (<1%). The frequency of the rtA181T mutation significantly increased after the second year, with a 50% increase during follow-up treatment. The frequency of the rtI224V mutation increased slowly. The frequency of the rtN248H mutation was high (70%). The fluctuations in other resistance mutations were relatively minor.

Figure 7. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV monotherapy in patient 7

A. The serum HBV DNA level fluctuated around 10⁴ to 10⁵ IU/ml during long-term ADV monotherapy. B. At baseline, this patient harbored the rtA181T (1.2%), rtN236T (1.5%), and rtV214A (1%) substitutions, and other resistance mutations were present at low levels (<1%). The frequencies of the rtA181V and rtN236T mutations increased slowly (ranging from 1.22% to 4.78% and 1.51% to 3.57%) after long-term treatment. The rtR153Q mutation was always present at a high frequency (70%). The frequency of the rtN139K mutation slowly increased after one year of treatment. The frequencies of the rtI224V, rtS223A, rtD134E mutations slowly increased after 4 years of treatment. The fluctuations in the frequency of other resistance mutations were relatively minor.

Figure 8. The dynamic changes in HBV DNA levels and resistance-associated variants in the RT domains during ADV monotherapy in patient 8

A. The patient underwent long-term ADV monotherapy, then experienced virological breakthrough after 3 years of treatment. The viral load remained at 10⁶ IU/ml in follow-up therapy. B. This patient harbored the rtA181T (1.3%) and rtN236T (1.6%) substitutions at baseline, and other resistance mutations were present at low levels (<1%). It is noteworthy that the frequency of the rtN236T mutation began to increase significantly after two years of ADV monotherapy and was present in up to 50% of circulating viruses during follow-up treatment. The frequency of the rtA181T mutation was more than 10% after 4 years of treatment. The frequency of the rtI224V mutation was always high (70~80%).