Effect of switching from tenofovir disoproxil fumarate to tenofovir alafenamide on lipid profiles in patients with hepatitis B

For long-term treatment of hepatitis B virus (HBV) infection, switching from tenofovir-disoproxil-fumarate (TDF) to tenofovir-alafenamide (TAF) may prevent renal dysfunction and bone loss. However, the precise effects of this switch on the blood lipid profile remain to be clarified. This is an important issue as TDF is known to have effects on both low- and high-density lipids. Therefore, our retrospective multi-center study aimed to evaluate the effects of switching from TDF to TAF on the lipid profile of patients with HBV infection. Samples were obtained prior to the switch from TDF to TAF and at 6–12 months after TAF initiation. In some cases, additional samples obtained pre- and post-TDF administration were available for analysis. Serum cholesterol levels, including oxidized-low-density lipoprotein (LDL) and non-high-density lipoprotein-cholesterol (HDL-c), and the rate of dyslipidemia, according to the NCEP-ATP III lipid risk classification, were analyzed. The data from 69 patients were analyzed, including 33 patients with pre- and post-TDF-initiation serum samples. Total cholesterol (T-chol), HDL-c, LDL-c, non-HDL-c, and oxidized LDL levels increased significantly after switching to TAF. With regard to sequential changes pre- to post-TAF, TDF was associated with significantly lower serum T-chol, HDL-c, and oxidized LDL-c levels, with T-chol, HDL-c, LDL-c, and oxidized LDL-c levels increasing significantly after the switch. The switch from TDF to TAF was also associated with an increase in the rate of dyslipidemia, from 33% to 39%, with an increase in the rate of severe dyslipidemia of 1.4% and 5.8%, based on T-chol and LDL-c levels. Of note, no cases of severe dyslipidemia were detected pre-TAF treatment. As oxidized LDL-c and non-HDL-c are strongly associated with atherosclerosis development, careful monitoring of lipid is needed after switching from TDF to TAF in this clinical population.


Introduction
Hepatitis B virus (HBV) infection is a major cause of hepatocellular carcinoma (HCC) and liver cirrhosis and accounts for approximately 1 million worldwide deaths annually [1][2][3]. Thus, effective therapeutic strategies to manage HBV infection are needed. Currently, interferon and nucleos(t)ide analogs (NAs) are the most common therapies used for patients with an HBV infection, with good effectiveness and safety profiles having been achieved. However, NAs can require consecutive, and even indefinite, cycles of administration [4]. Lamivudine, adefovir dipivoxil, entecavir (ETV), tenofovir-disoproxil-fumarate (TDF), and tenofovir alafenamide (TAF) are approved NAs for chronic hepatitis B and are considered as first-line NAs for HBV infection [4][5][6][7]. Both TDF and TAF are pro-drugs of tenofovir [8]. However, longterm administration of TDF can cause renal dysfunction and bone loss [9][10][11]. By contrast, TAF was designed to have greater stability than TDF, such that the active metabolite of tenofovir diphosphate is delivered to the liver more efficiently. As such, the required dose of TAF (25 mg once a day) is lower than the required dose of TDF (300 mg once a day), resulting in less kidney and bone toxicity [8]. Switching from TDF to TAF has been conducted in general clinical practice for patients who are at risk of renal dysfunction and bone loss.
The positive intra-and extrahepatic effects of TDF are important to note, however, when considering a switch from TDF to TAF. Murata et al. [12] clearly demonstrated that TDF promotes the release of interferon-λ3 by gastrointestinal cells, which could affect serum hepatitis B surface antigen (HBsAg) levels. In addition, we and other groups have shown that, in patients with HBV and/or human immunodeficiency virus (HIV) infection, TDF reduces the serum cholesterol level through the upregulation of the CD36/PPAR-alpha axis [13][14][15]. Previous research has shown that switching from TDF to TAF increases serum cholesterol levels in patients with HIV infection [16]. However, it is important to note that TDF may increase both low-density lipoprotein cholesterol (LDL-c) and high-density lipoprotein cholesterol (HDL-c) levels, wherein an increase in LDL-c can promote development of atherosclerosis, while an increase in HDL-c can provide a protective effect [17]. Thus, the actual effect of switching from TDF to TAF on the development of atherosclerosis is unclear.
We have recently reported that TDF could decrease serum oxidized LDL-c [15], which is strongly associated with the development of atherosclerosis [18,19]. However, the kinetics of oxidized LDL-c after switching from TDF to TAF has not yet been clearly established. Therefore, this study aimed to evaluate the effect of switching from TDF to TAF on the lipid profile of patients with a chronic HBV infection. Additionally, we analyzed sequential changes in lipid profile components, including oxidized LDL-c, from baseline to TDF administration, after switching from TDF to TAF, and post-TAF administration for patients who had proper clinical information and preserved serum samples.

Patients and clinical study design
This was a retrospective multicenter study of patients with chronic hepatitis B who switched from TDF to TAF, recruited between 2016 and 2020 from hospitals in the NORTE Study Group, who performed clinical studies of liver diseases [20][21][22][23][24][25]. Included were patients with an HBV infection who were �20 years of age, had switched from TDF to TAF for the management of their HBV infection and received TAF for >6 months, and for whom the required information was available in their medical record and preserved serum samples at baseline and 6-12 months after the initiation of TAF were available. Patients with additional clinical information and preserved serum samples at all time points (pre-TDF administration, post-TDF administration [6-12 months], pre-TAF administration, and post-TAF administration [6-12 months]) were included in detailed analyses of sequential changes in the lipid profile. Patients were excluded if they had a history of liver disease other than chronic hepatitis B, uncontrolled malignancy, including HCC, a history of treatment for dyslipidemia, had received prophylactic NA for HBV reactivation, or declined to participate in the study.
Clinical information was collected, including laboratory data, HBV infectious status, and concomitant drug information, in addition to preserved serum samples. Changes in cholesterol levels were analyzed in preserved serum samples. TAF (25 mg) or TDF (300 mg) was administered once daily according to the manufacturer's recommendation.

Statement of ethics
The study protocol conformed to the ethical guidelines of the Declaration of Helsinki and was approved by the ethics committee of Hokkaido University Hospital (Clinical research number: IRB 017-0387). All enrolled patients provided written informed consent to participate in this study or did not decline to participate in this study. All data were fully anonymized before we accessed them, and the ethics committee specifically approved non-decline of being included in the study in lieu of written informed consent for some patients.

Serum HBV markers and lipid profile analysis
Serum total cholesterol (T-chol), HDL-c, LDL-c, and oxidized LDL levels were analyzed by enzyme-linked immunosorbent assays (SRL, Tokyo, Japan), according to manufacturer's instructions, as previously described [15]. Serum HBsAg and HBV-DNA levels were analyzed by chemiluminescence immunoassay (Architect HBsAg-QT assay; Abbott Laboratory, Tokyo, Japan) and a real-time TaqMan PCR assay (Cobas AmpliPrep/Cobas TaqMan HBV Test, v2.0) [10,26]. The change in the rate of dyslipidemia was evaluated based on the National Cholesterol Education Program-Adult Panel III (NCEP-ATP III) lipid risk classification [27].

Statistical analysis
Categorical variables were analyzed using the χ2 test. Continuous variables were analyzed using the paired t-test or Mann-Whitney U test. Differences among three or more groups were evaluated using Friedman's test and Dunn's multiple comparisons tests. Differences in the proportion of NCEP-ATP III classification-based dyslipidemia between pre-TAF and post-TAF samples were assessed using the extended McNemar's test. To analyze changes in the rate of dyslipidemia across time points, multiple logistic regression was used. All P-values were two-tailed, and the level of significance was set at P < 0.05. Statistical analyses were conducted using Prism 7.03 (GraphPad Software, Inc., La Jolla, CA, USA).

Patient characteristics
A total of 99 patients with chronic hepatitis B were switched from TDF to TAF between 2016 and 2020. Among these, 30 patients were excluded owing to a lack of preserved serum samples, insufficient data, and/or treatment for dyslipidemia (Fig 1), with the remaining 69 patients enrolled (Table 1). Preserved serum samples obtained pre-TDF initiation and post-TDF administration (6-12 months after TDF initiation) were available from 33 of these 69 patients for analyses of sequential changes in lipid profiles and clinical data (Table 1).

Lipid profile changes after switching from TDF to TAF in patients with chronic hepatitis B
As shown in Fig 2A, the switch from TDF to TAF significantly increases the levels of T-chol (P < 0.01), HDL-c (P < 0.05), LDL-c (P < 0.01), non-HDL-c (P < 0.01), and oxidized LDL (P< 0.01) at 6-12 months after the initiation of TAF. Similar tendencies were observed both in patients with liver cirrhosis and chronic hepatitis (S1 Fig). The sequential analysis of samples from the 33 patients with sufficient serum samples revealed that TDF administration significantly lowered serum T-chol, HDL-c, and oxidized LDL-c levels. The switch from TDF to TAF significantly increased T-chol, LDL-c, and oxidized LDL-c levels ( Fig 2B).

Changes in the rate of dyslipidemia after switching from TDF to TAF
We analyzed the change in the proportion of patients with dyslipidemia according to the NCE-P-ATP III lipid risk classification before and after the switch from TDF to TAF. As shown in Fig 3A, the rate of dyslipidemia, as evaluated by T-chol and LDL-C, increased significantly after switching from TDF to TAF. The rate of dyslipidemia, as determined by HDL-c, decreased significantly after switching from TDF to TAF. Overall, the proportion of patients with dyslipidemia based on any indicator increased from 33% to 39% by switching from TDF to TAF. In addition, the rate of severe dyslipidemia increased to 1.4% to 5.8%, based on T-chol and LDL-c levels, respectively, after the switch from TDF to TAF, with no case of severe dyslipidemia identified pre-TAF treatment. Furthermore, the sequential analysis indicated a significant difference in the proportion of patients with dyslipidemia (severe dyslipidemia plus dyslipidemia), based on the LDL-c level and the NCEP-ATP III lipid risk classification, from pre-TDF to post-TDF (6-12 months) administration (P = 0.001; Fig 3B).

Changes in renal function, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and HBV infection status after switching from TDF to TAF
As shown in Figs 4A and 5A, after switching from TDF to TAF, the estimated glomerular filtration rate (eGFR), AST, and ALT levels at 6-12 months after TAF initiation and those prior to TAF administration were not different. Among the 33 patients included in the sequential analysis, the mean eGFR, AST, and ALT levels decreased significantly from pre-to post-TDF administration and were not affected by switching from TDF to TAF (Figs 4B and 5B). As shown in Table 2, HBsAg levels decreased significantly with both TDF and TAF administration.

PLOS ONE
Effect of TDF-to-TAF switching on lipids

PLOS ONE
Effect of TDF-to-TAF switching on lipids

Discussion
Our results indicate that switching from TDF to TAF significantly increased T-chol, LDL-c, HDL-c, oxidized LDL-c, non-HDL-c levels, as well as the rate of dyslipidemia, in patients with chronic hepatitis B. As oxidized LDL-c and non-HDL-c levels are strongly associated with the development of atherosclerosis [18,19,28], careful monitoring of these parameters is needed when switching from TDF to TAF. NAs are safe and effective for patients with HBV infection and thus are widely used, even for long-term treatment. In particular, ETV, TDF, and TAF are first-line NAs for the management of HBV infection [5]. However, long-term TDF administration is associated with renal dysfunction and bone loss [9][10][11]. For the treatment of HBV infection over a 48-week period, the effect of TAF is non-inferior to that of TDF, with improved bone and renal safety compared with that of TDF [29]. Over an extended duration of 96 weeks, TAF is as effective as TDF, with continued improved renal and bone safety [8]. Thus, patients with a risk of renal and bone diseases may benefit from switching from TDF to ETV or TAF, according to the European Association for the Study of the Liver guideline [5]. In our study, eGFR did not change after switching from TDF to TAF (Fig 4A) although gradually declined during TDF treatment (Fig 4B).
In patients with HIV and/or HBV infection, previous reports have shown that TDF administration lowers T-Chol, LDL-c, and HDL-c levels simultaneously [13][14][15]. Consistent with these previous reports, in our study, cholesterol levels decreased significantly during TDF administration (Fig 2B). However, other reports have shown that switching from TDF to TAF increased levels of cholesterols, including T-chol, LDL-c, and HDL-c, in patients with HIV infection [16,30,31]. As HDL-c and LDL-c have opposite effects on the development of atherosclerosis, the effects of changing both serum parameters are not clear. Recently, we reported that TDF for patients with HBV decreases serum cholesterol levels, including non-HDL and oxidized LDL, which are strongly associated with atherosclerosis [15,18,19,28]. Macrophages phagocytize oxidized LDL-c and are subsequently transformed into foam cells, resulting in atherosclerosis plaque formation [32]. Blocking the uptake of oxidized LDL-c by macrophages inhibits the development of atherosclerosis [19]. Thus, the reductions in serum oxidized LDLc and non-HDL levels by TDF are believed to reduce the risk of atherosclerosis. In our study, we showed that switching from TDF to TAF significantly increased levels of non-HDL-c and oxidized-LDL. Therefore, the risk of development of atherosclerosis might increase after a switch from TDF to TAF.
Both TDF and TAF are pro-drugs of tenofovir [8], and the mechanisms by which switching from TDF to TAF increases non-HDL-c and oxidized-LDL levels are therefore unclear. The difference in the dosage of TAF (25 mg once a day) and TDF (300 mg once a day) might explain the observed differential effect on cholesterol levels. The lower bone and kidney toxicity of TAF has been attributed to the lower required dosage of TAF (25 mg once a day) than that of TDF (300 mg once a day) [8].

PLOS ONE
Effect of TDF-to-TAF switching on lipids oxidized LDL-c [35], and HDL-c [36]. Furthermore, we revealed that the effect of TDF in reducing supernatant cholesterol was dose-dependent; therefore, the attenuation of the effect on cholesterol levels might be related to the difference in required dosage between TAF and TDF.
Moreover, the rate of dyslipidemia, according to LDL-c and T-chol levels, as defined by the NCEP-ATP III lipid risk classification, increased after switching from TDF to TAF (from 0%

PLOS ONE
Effect of TDF-to-TAF switching on lipids with TDF administration to 14.4% for T-chol and from 13.0% to 31.4% for LDL-c). The rate of dyslipidemia determined by HDL-c levels decreased (from 23.2% with TDF administration to 7.2%). Overall, the rate of dyslipidemia based on any cholesterol increased from 33% to 39% after switching from TDF to TAF. Additionally, after switching, 1.4% and 5.8% of patients showed severe dyslipidemia based on T-chol and LDL-c, respectively, which was not observed pre-TAF treatment. Based on these findings, switching from TDF to TAF could have negative effects on the lipid profile, both qualitatively and quantitatively. Thus, this switch should be considered with caution.
Increased cholesterol levels after switching from TDF to TAF might affect intra-and extrahepatic disorders in patients with HBV infection. A high T-chol level in men with HBV infection is associated with an increased risk of HCC [37]. Thus, in such patients, the increase in cholesterol levels by switching from TDF to TAF requires careful monitoring for HCC. Additionally, TDF reduces the incidence of heart failure, which is sometimes associated with atherosclerosis [38] and reduces the lower common carotid intima-media thickness in patients with HIV infection [39]. Thus, after switching from TDF to TAF, the incidence of atherosclerosis-related diseases also requires careful attention.
The limitations of our study need to be acknowledged. First, it was a retrospective study with a relatively limited sample size. Furthermore, we analyzed cholesterol levels in preserved serum samples, which were collected regardless of food intake, preventing accurate assessments of changes in serum triglyceride levels. Additionally, data related to physical activity and energy intake were lacking, which could affect lipid metabolism, and data of cholesterols that changed after TAF administration in nucleotide analogue-naive HBV patients were not obtained. Therefore, larger prospective studies are required to validate our results.

Conclusion
To the best of our knowledge, we provide the first evidence that switching from TDF to TAF not only increases T-chol, LDL-c, and HDL-c levels but also increases oxidized LDL-c and non-HDL-c levels in patients with chronic hepatitis B. In addition, switching from TDF to TAF increased the proportion of patients with dyslipidemia. As oxidized LDL-c and non-HDL-c are strongly associated with atherosclerosis, these changes should be carefully assessed when switching from TDF to TAF.