Comparative effectiveness and safety of tenofovir alafenamide versus tenofovir disoproxil fumarate in patients with HBV-associated decompensated cirrhosis: insights from a retrospective analysis

Abstract

Background

HBV infection is a global health issue. Given the severity of decompensated cirrhosis, selecting the most appropriate first-line antiviral medication is crucial for optimizing treatment outcomes and providing the best possible disease control. This retrospective real-world study aimed to evaluate the comparative effectiveness and safety profiles of tenofovir alafenamide (TAF) versus tenofovir disoproxil fumarate (TDF) in patients with hepatitis B virus (HBV)-related decompensated cirrhosis. We hypothesised that TAF is non-inferior to TDF for virologic response and superior for renal safety, with corresponding statistical tests to assess non-inferiority and superiority.

Methods

Patients receiving TAF or TDF were categorized into respective groups based on their treatment regimen. The study spanned 48 weeks, during which the impact on HBV replication, liver enzyme levels, and renal function was assessed. Propensity score matching (PSM) was utilized to minimize disparities between TAF and TDF cohorts.

Results

After 48-week treatment, the percentage of patients with detectable HBV DNA significantly decreased from 27.9 to 13.1% (p = 0.044). HBeAg positivity remained stable, increasing slightly from 27.9 to 31.15% (p = 0.887). Mean ALT levels decreased slightly from 26 U/L to 25 U/L, with 85.25% of patients achieving normal ALT levels by week 48. TAF was well-tolerated with no significant impact on renal function, as indicated by stable serum creatinine and eGFR levels (eGFR decline ≥ 25% occurred in 5.2%). After PSM (propensity score matching), both groups showed virologic suppression, but TAF demonstrated superior renal safety: serum creatinine remained stable in TAF (P = 0.921) versus a trend toward increase in TDF (P = 0.831); eGFR significantly declined in TDF (103.58 → 99.05 mL/min/1.73 m²) but remained stable in TAF (101.88 → 101.86 mL/min/1.73 m²); clinically relevant eGFR decline (≥ 25%) was lower with TAF (5.2% vs. 10.5% in TDF). TAF also showed a higher rate of ALT normalization (71.1% → 81.6%, P < 0.05) compared to stable rates in TDF (84%). For liver fibrosis markers, APRI scores decreased non-significantly in TAF (P = 0.365) but increased in TDF (P = 0.763), while FIB-4 scores slightly decreased in TAF (P = 0.527) versus a non-significant increase in TDF. Child-Pugh scores remained minimally changed in both groups.

Conclusion

TAF appears to offer similar virologic response rates to TDF, with potentially superior renal safety. This study provides valuable insights into the management of HBV-related decompensated cirrhosis, advocating for TAF’s consideration in treatment protocols.

Trial registration

All procedures were performed in accordance with the guidelines of the institutional ethics committee and adhered to the tenets of the Declaration of Helsinki. This study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University, China, which also waived the need for informed consent (approval number: KY2024-101). The registration date is April 24, 2024.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-11473-6.

Background

Hepatitis B virus (HBV) infection is a pervasive and devastating global health issue, with a profound impact on liver health [1]. It is a major etiological agent in the development of acute and chronic liver diseases, which often progress to liver fibrosis, and eventually to the more severe stages of liver cirrhosis and hepatocellular carcinoma (HCC). The journey from initial infection to these advanced liver pathologies is a testament to HBV’s insidious nature and its ability to evade the host’s immune responses. For many individuals, the diagnosis of HBV-related cirrhosis comes at a decompensated stage, complicating treatment strategies and reducing the likelihood of favorable outcomes.

In the context of liver cirrhosis, the term “decompensated” signifies a critical and often dire stage in the disease’s progression. Decompensated cirrhosis is characterized by the failure of the liver to perform its essential functions, leading to a cascade of complications such as ascites, jaundice, encephalopathy, and gastrointestinal bleeding—each of which poses a significant threat to the patient’s life. The onset of decompensation typically marks a transition from a more manageable compensated phase, where the liver is still able to fulfill its roles despite underlying damage [2].

Given the severity of decompensated cirrhosis, selecting the most appropriate first-line antiviral medication is crucial for optimizing treatment outcomes and providing the best possible disease control [3]. Despite the availability of effective antiviral medications like Entecavir (ETV), Tenofovir (TDF), and Tenofovir Alafenamide (TAF), the specific treatment of decompensated cirrhosis due to HBV remains a clinical challenge. These drugs are recommended as first-line therapies for chronic HBV infection because of their potent antiviral activity and high genetic barrier to resistance [4, 5]. No studies have compared the efficacy and safety of these drugs in decompensated cirrhosis. However, in these drugs, while TDF and TAF belong to the same class of drugs, research indicates that they differ in terms of bone and renal safety profiles. Most of these studies, however, have focused on the general population. In the vulnerable population of patients with HBV-related decompensated cirrhosis, the question of how to precisely choose the most suitable first-line treatment to better control disease progression becomes even more critical. Despite the importance of this issue, the potential differences in the efficacy and safety of TDF and TAF in treating decompensated cirrhosis remain largely unexplored. The choice between TDF and TAF could be pivotal for patients with decompensated cirrhosis, potentially determining the trajectory of their disease. For clinicians, understanding the nuanced differences between these antivirals is essential in crafting a treatment regimen that not only suppresses viral replication but also mitigates the risks of further renal or bone complications, which could exacerbate the already delicate condition of these patients.

Based on previous research, we hypothesised that TAF is non-inferior to TDF for virologic response and superior for renal safety. To test these hypotheses, our research team has embarked on a 48-week cohort study to assess the efficacy and safety of TAF in patients with HBV-related decompensated cirrhosis, as well as its comparison with TDF. Our primary objective in this pivotal research endeavor is to establish whether TAF exhibits equivalent therapeutic efficacy and a comparable safety profile when administered during the decompensated phase of cirrhosis. Furthermore, the comparative analysis between TAF and TDF is designed to provide insights into the relative merits of these two first-line antiviral agents in the context of decompensated cirrhosis. This comparative evaluation is crucial, given that while both drugs have demonstrated potent antiviral activity, their effects on the liver’s structural integrity and the overall clinical outcomes in patients with advanced cirrhosis have yet to be thoroughly explored. In conclusion, the selection of the optimal first-line antiviral therapy for patients with decompensated cirrhosis is of paramount importance. Our study seeks to fill a critical gap in the current understanding of TAF and TDF’s roles in managing HBV-related decompensated cirrhosis, offering clinicians the evidence needed to make informed decisions that could significantly impact patient outcomes and improve the quality of life for those grappling with this life-threatening condition.

Materials and methods

Patients and methods

In a retrospective analysis, we meticulously reviewed the medical records of patients with HBV-related decompensated cirrhosis who initiated antiviral therapy (TAF 25 mg/day or TDF 300 mg/day) at the First Affiliated Hospital of Wenzhou Medical University. Decompensation was defined by at least 1 event: ascites (ultrasound/clinical), hepatic encephalopathy (West-Haven criteria), hepatorenal syndrome, or variceal bleeding (endoscopy). Cases were identified via ICD-10 codes and chart validation by two hepatologists. The study period spanned six years, from January 1, 2018, to December 31, 2023. To ensure robust study outcomes, only those with a minimum follow-up duration of 48 weeks were included. In the end, a total of 89 patients treated with TAF and 107 patients treated with TDF were included in this study, of whom 65 were excluded due to incomplete data (Fig. S1).

To meticulously align the cohorts for meaningful comparison, we employed the propensity score matching method (PSM). This statistical technique adeptly minimized disparities between the TAF and TDF groups across various parameters, including age, gender, HBeAg status, baseline HBV DNA levels, serum creatinine and ALT levels. PSM was performed 1:1 using a logistic regression model (caliper width = 0.02). Following matching, 38 matched pairs were enrolled in the final analysis. Standardized Mean Differences (SMD) for all covariates are now reported in Table 1.

Table 1.

Baseline characteristics before and after PSM in TAF/TDF group

	Before PSM				After PSM
	TAF (n = 61)	TDF (n = 70)	P value	SMD	TAF (n = 38)	TDF (n = 38)	P value	SMD
Age, years	52.31 ± 9.48	55.21 ± 11.44	0.12	−0.2	53.05 ± 8.85	52.39 ± 11.18	0.78	0.06
Male, n (%)	47/61(77.0%)	55/70(78.6%)	0.834	-	29/38(80.6%)	27/38(75%)	0.57	-
HBV DNA-positive	17/61(27.9%)	40/70(57.14%)	< 0.001	-	14/38(36.8%)	20/38(52.6%)	0.166	-
HBeAg-positive	17/61(27.9%)	19/70(27.14%)	0.926	-	11/38(28.9%)	11/38(28.9%)	-	-
ALT, U/L	32.08 ± 20.67	52.76 ± 57.87	0.009	−0.53	36.66 ± 22.33	30.32 ± 17.75	0.18	0.01
Ratio of normal ALT	48/61(78.69%)	46/70(71%)	0.1	-	27/38(71.1%)	32/38(84.2%)	0.169	-
eGFR, mL/min/1.73 m2	94.63 ± 25.21	106.15 ± 18.40	0.003	0.55	101.88 ± 18.82	103.58 ± 20.97	0.71	−0.08
Scr, µmol/L	82.08 ± 34.96	67.94 ± 16.30	0.003	0.53	71.80 ± 20.98	71 ± 19.42	0.86	0.04

HBV DNA: negative (< 20 IU/mL); ALT: ≤ 40 U/L; eGFR: ≥ 90 mL/min/1.73 m² (CKD-EPI equation); Scr: 44–97 µmol/L

Inclusion criteria and exclusion criteria

The study was predicated on the definition of decompensated cirrhosis, marked by the emergence of significant clinical manifestations such as ascites, hepatorenal syndrome, hepatic encephalopathy, and variceal hemorrhage. Inclusion criteria: age > 18 years, a diagnosis of decompensated cirrhosis secondary to chronic hepatitis B, and ongoing antiviral therapy with either TAF or TDF. Exclusion criteria: concomitant hepatocellular carcinoma, HCV or HIV coinfection, autoimmune liver disease or prior transplant.The study protocol, encompassing the use of electronic medical record systems for data collection, was dutifully approved by the Ethical Committee of the First Affiliated Hospital of Wenzhou Medical University. Given the retrospective nature of the study, the requirement for informed consent was waived. Ethical Considerations: KY2024-101.

Statistical analysis

Data were rigorously analyzed, with continuous variables of normal distribution presented as mean ± standard deviation (SD) and compared using t-tests. Non-normally distributed continuous variables are represented by the median (interquartile range). Categorical variables were articulated as numbers (percentages) and juxtaposed using chi-squared tests or Fisher’s exact tests, as appropriate. A p-value of < 0.05 was set as the threshold for statistical significance. All analyses were conducted with precision using IBM SPSS software, version 27.0.

Results

Baseline characteristic

The demographic and clinical characteristics of the patients at the onset of the study are detailed in Table 2. In the TAF group, the average age was 52.31 years, with a standard deviation of 9.48 years. A majority, 77.0% (47 out of 61), were male. HBeAg positivity was observed in 27.9% (17 out of 61) of the patients. Among these patients, 27.9% of them had detectable HBV DNA at baseline. The initial ALT levels averaged at 32.08 ± 20.67 U/L, with a normal ALT ratio observed in 78.69% (48 out of 61) of the patients.

Table 2.

Baseline characteristic of TAF group

Age, years	52.31 ± 9.48
Male, n (%)	47/61(77.0%)
HBV DNA-positive	17/61(27.9%)
HBeAg-positive	17/61(27.9%)
ALT, U/L	32.08 ± 20.67
Ratio of normal ALT	48/61(78.69%)
eGFR, mL/min/1.73 m2	94.63 ± 25.21
Scr, µmol/L	82.08 ± 34.96
Age, years	52.31 ± 9.48

HBV DNA: negative (< 20 IU/mL); ALT: ≤ 40 U/L; eGFR: ≥ 90 mL/min/1.73 m² (CKD-EPI equation); Scr: 44–97 µmol/L

TAF effectiveness

Loss of HBeAg and virologic response

In our analysis, a notable majority of patients within the TAF group had previously received nucleos(t)ide analog therapy. Consequently, only 27.9% of these patients exhibited detectable HBV DNA levels at the commencement of the study. Following 48 weeks of treatment with TAF, this figure significantly decreased to 13.1% (8/61). The reduction in HBV DNA from baseline to week 48 was found to be statistically significant (p = 0.044). HBV-DNA declines (log-IU/mL) for all patients were also reported (Fig. S2). In baseline viremic patients (n = 17/61), VR rates were 76.5%(P < 0.001). All the above data demonstrate that TAF achieves a robust virologic response (Fig. 1A).

Fig. 1.

Effects (A to D) and safety (E, F) of TAF after 48-week treatment, and Evolution of Child-Pugh, FIB-4, and APRI Scores (G)

Regarding HBeAg status, the proportion of patients who tested positive for HBeAg was relatively stable, with 27.9% (17 out of 61) at the outset and 31.15% (19 out of 61) at the 48-week mark. This difference was not statistically significant (P = 0.887), suggesting that while there was a slight increase in HBeAg positivity, it did not reach a level of clinical relevance (Fig. 1B).

ALT normalization ratio

After TAF treatment, patients exhibited a trend of decreasing mean ALT levels over the 48-week treatment period. Specifically, the mean ALT levels were recorded as 26(19) U/L at baseline and reduced to 25(18) U/L by week 48 (Fig. 1C). Although there was a decline, the difference did not reach statistical significance when compared to the initial measurements (p = 0.332). Nonetheless, by week 48, an impressive 85.25% of patients (52 out of 61) achieved normalization of ALT levels (Fig. 1D).

Safety profiles of TAF

During the 48-week follow-up period, we meticulously monitored and compared the mean changes in serum creatinine and estimated glomerular filtration rate (eGFR) against their respective baseline values within the TAF group. The serum creatinine levels (Fig. 1E) exhibited a non-significant decrease by a mean of 73(31.5) µmol/L at week 48 (p = 0.583). Similarly, the eGFR(mL/min/1.73m²) (Fig. 1F) demonstrated no significant deviation from baseline throughout the study period (P = 0.774). 5.7% of them experienced an eGFR decline exceeding 25%. These findings collectively demonstrate that TAF was well-tolerated with no discernible impact on renal function.

Child-Pugh, FIB-4, and APRI scores

We conducted a comprehensive assessment of liver fibrosis by comparing the AST to Platelet Ratio Index (APRI) scores, FIB-4 scores and Child-Pugh scores before and after the treatment intervention (Fig. 1G). At the outset of the study, the APRI scores were recorded with a median value of 0.89 (1.22), which remained stable by week 48 with a median of 0.80 (0.93). Similarly, the FIB-4 scores showed a slight decrease from a baseline median of 3.55 (4.24) to 3.45 (3.12) at week 48. The Child-Pugh scores also demonstrated a minimal change, with median values shifting from 6 (2) at baseline to 6 (1) by week 48. These findings indicated that inn patients with decompensated cirrhosis, the improvements in hepatic fibrosis and liver function observed with TAF were less pronounced.

Effectiveness and safety of TAF versus TDF

Patient cohorts post-propensity score matching (PSM)

Significant imbalances were observed in TAF and TDF groups: patients assigned to TDF were older (55.21 ± 11.44 vs. 52.31 ± 9.48 years; P = 0.12, SMD = − 0.20), more often HBV DNA-positive (57.1% vs. 27.9%; P < 0.001), had higher ALT levels (52.76 ± 57.87 vs. 32.08 ± 20.67 U/L; P = 0.009, SMD = − 0.53), higher eGFR (106.15 ± 18.40 vs. 94.63 ± 25.21 mL/min/1.73 m²; P = 0.003, SMD = 0.55) and lower serum creatinine (67.94 ± 16.30 vs. 82.08 ± 34.96 µmol/L; P = 0.003, SMD = 0.53).

To mitigate these baseline imbalances, we conducted a propensity score matching. After matching, we successfully created 38 matched pairs between the TAF and TDF cohorts). After PSM, all baseline variables were well balanced. Mean age was 53.05 ± 8.85 years in the TAF group and 52.39 ± 11.18 years in the TDF group (P = 0.78, SMD = 0.06). Proportions of males, HBV DNA-positive patients, and HBeAg-positive patients were comparable (all P ≥ 0.05). Laboratory indices also showed excellent balance: ALT 36.66 ± 22.33 vs. 30.32 ± 17.75 U/L (P = 0.18, SMD = 0.01), eGFR 101.88 ± 18.82 vs. 103.58 ± 20.97 mL/min/1.73 m² (P = 0.71, SMD = − 0.08), and serum creatinine 71.80 ± 20.98 vs. 71.00 ± 19.42 µmol/L (P = 0.86, SMD = 0.04). All post-match SMDs were < 0.1, confirming successful matching.

Renal function assessment

Over the 48-week treatment period, there was a noted increase in serum creatinine levels in both the TAF and TDF groups (Fig. 2A; Table 3). In the TAF group, serum creatinine levels rose from a baseline of 71.80 ± 20.98µmol/L to 72.56 ± 28.77µmol/L by week 48; this change was not statistically significant (P = 0.921). In contrast, the TDF group experienced aincrease from 71 ± 19.42µmol/L at baseline to 71.92 ± 18.09µmol/L at week 48, which was not statistically significant(P = 0.831).

Fig. 2.

Effectiveness and safety of TAF versus TDF

Table 3.

Key laboratory parameters at baseline (Week 0) and week 48 in the propensity-score-matched TAF and TDF groups

	TAF(n = 38)			TDF(n = 38)
	0w	48w	P value	0w	48w	P value
HBV DNA-positive	14(36.8%)	8(21.1%)	0.129	20(52.6%)	4(10.5%)	< 0.001
HBeAg-positive	11(28.9%)	11(28.9%)	-	11(28.9%)	11(28.9%)	-
ALT, U/L	36.66 ± 22.33	31 ± 18.38	0.261	30.32 ± 17.75	29.42 ± 12.47	0.8
Ratio of normal ALT	27(71.1%)	31	0.28	32/38(84.2%)	31	0.761
eGFR, mL/min/1.73 m2	101.88 ± 18.82	101.86 ± 22.24	0.928	103.58 ± 20.97	99.05 ± 20.33	0.343
Scr, µmol/L	71.80 ± 20.98	72.56 ± 28.77	0.921	71 ± 19.42	71.92 ± 18.09	0.831

HBV DNA: negative (< 20 IU/mL); ALT: ≤ 40 U/L; eGFR: ≥ 90 mL/min/1.73 m² (CKD-EPI equation); Scr: 44–97 µmol/L

Regarding eGFR, the figure in TAF group remained virtually unchanged from baseline to week 48 (101.88 ± 18.82 vs. 101.86 ± 22.24 mL/min/1.73 m²). In contrast, the TDF group exhibited a discernible downward trend, with mean eGFR decreasing from 103.58 ± 20.97 to 99.05 ± 20.33 mL/min/1.73 m², reflecting a modest but consistent decline in renal function that was not observed with TAF. We also find that eGFR decline ≥ 25% occurred in 5.2% (TAF) vs. 10.5% (TDF; p = 0.39).

ALT levels and ALT normalization

Table 3 shows a clear ALT benefit favouring TAF. Mean ALT fell from 36.66 ± 22.33 U/L at baseline to 31.0 ± 18.38 U/L at week 48, indicating a sustained biochemical improvement. Consequently, the proportion of TAF-treated patients with normal ALT rose from 71.1% (27/38) to 81.6% (31/38).

In contrast, the TDF group displayed minimal change: baseline ALT was 30.32 ± 17.75 U/L and remained almost identical at week 48 (29.42 ± 12.47 U/L), with the normalisation rate already high and unchanged at 84%.

Although the findings above numerically favoured TAF, longer follow-up and adjustment for baseline ALT, BMI and fibrosis stage are needed to clarify any differential hepatic benefit.

Liver fibrosis indicators

In the portion of our study that concentrated on treatment outcomes, the TAF group showed a decline in APRI scores after 48 weeks of treatment, yet this decline was not statistically significant (P = 0.365, Fig. 3A). In contrast, the TDF group exhibited a increase in APRI scores during the same timeframe (P = 0.763).

Fig. 3.

Liver Fibrosis Indicators of patients in TAF or TDF group (after PSM)

Upon analyzing the FIB-4 scores (Fig. 3B), the baseline value for the TAF group was 5.22 ± 5.45, which slightly decreased by week 48(4.54 ± 4.28), though this change was also not statistically significant (P = 0.527). Conversely, the TDF group, starting with a higher baseline FIB-4 score of 5.45 ± 4.04, displayed a slightly upward trend. Child-Pugh scores exhibited comparable trajectories in both arms, with no appreciable divergence over the observation period.(Fig. 3C).

Discussion

In this retrospective, real-world study, we categorized patients into TAF or TDF groups based on their treatment regimen. Over the 48-week treatment period, the results demonstrated that, in HBV-related decompensated cirrhosis, TAF was highly effective in suppressing HBV replication in the blood. Meanwhile, treatment with TAF was associated with ALT normalization, HBeAg loss, and stable renal function without discernible nephrotoxicity. The present study also offers valuable insights into the comparative effectiveness and safety profiles of TAF and TDF in patients with HBV-related decompensated cirrhosis. The propensity score matching method (PSM) was used to reduce heterogeneity between the TAF and TDF cohorts, with respect to age, gender, HBeAg status, baseline HBV DNA levels, serum creatinine and ALT levels.

The high virologic response (VR) rates observed in the TAF group at week 48 are encouraging, considering the advanced liver disease in this patient cohort. VR is a critical endpoint during chronic hepatitis B (CHB) antiviral treatment, as it can lead to improved liver inflammation, histological fibrosis, and better clinical outcomes. Previous studies, such as Kim’s research [6], have emphasized the importance of persistently detectable HBV DNA levels in increasing the risk of hepatocellular carcinoma (HCC), especially in patients with cirrhosis. Jiwon Yang et al. retrospectively analyzed six hundred twenty-seven untreated patients with chronic hepatitis B related compensated cirrhosis and concluded that those with compensated cirrhosis and LLV(Lower Limit of Virus) had a significantly higher risk of HCC [7]. Maria Buti et al. demonstrated non-inferiority of TAF with TDF by a randomized, double-blind non-inferior trial in both HBeAg-positive and HBeAg-negative patients receiving TAF or TDF therapy [8]. Our study demonstrated that TAF achieved a robust virologic response, manifesting not only as a higher rate of viral suppression at week 48 but also as the achievement of sustained virologic response in the majority of patients who were viremic at baseline. However, there was no significant change in the HBeAg seroconversion rate at week 48. The reasons may be as follows: (1) HBeAg seroconversion typically requires extended therapy (> 2 years) even in compensated patients. Our 48-week endpoint aligns with phase 3 trial data showing HBeAg loss rates of 10–22% at 1 year, suggesting insufficient time for immune reconstitution. (2) Decompensated cirrhosis is characterized by immune exhaustion and impaired HBeAg seroconversion mechanisms. Future studies with extended follow-up should clarify whether prolonged TAF therapy can overcome these barriers.

The trend towards ALT normalization in both groups is a positive indicator of hepatic recovery. As confirmed by previous studies, TAF was acknowledged to be superior in ALT normalization [9, 10]. In our study, the TAF group and TDF group both showed a higher rate of ALT normalization at week 48, this is particularly relevant for patients with decompensated cirrhosis, where liver function preservation is crucial. The results showed that ALT levels in the TAF group declined after treatment, whereas they remained essentially unchanged in the TDF group; moreover, the rate of ALT normalization at week 48 was higher in the TAF group. These findings may suggest an advantage of TAF in improving hepatic function. However, many factors can influence ALT, and because this study was retrospective we were unable to collect data on potential confounders such as concomitant fatty liver disease, use of herbal medicines, or alcohol consumption.

The non-invasive liver fibrosis markers APRI and FIB-4 have been well-established for diagnosing chronic liver disease [11, 12]. Zhou et al. observed significant declines of APRI and FIB-4 scores in patients who continued antiviral therapy, compared with those untreated and those stopping treatment [13]. In our study, after 48 weeks of treatment, there was a tendency for both APRI and FIB-4 scores to decrease in patients treated with TAF. However, this trend was not observed in the TDF group. Unfortunately, the changes in APRI and FIB-4 scores within the TAF group were not statistically significant. This is likely attributable to the limited sample size and insufficient follow-up duration. Long-term studies, such as one that observed significant declines in APRI and FIB-4 scores over a 5-year ETV treatment, suggest that longer-term therapy may be necessary to observe such changes [14].

The renal safety of TAF is particularly important given TDF’s known nephrotoxicity. Many studies have demonstrated that TAF has a favorable safety profile in the kidneys [15–17]. Switching to TAF led to improved kidney function, particularly in those with CKD [18]. The superior renal safety profile of TAF compared to TDF is likely due to its primary fecal excretion, with less than 1% excreted via the kidneys. Furthermore, TAF is a prodrug of TDF designed to be activated in the liver, achieving high concentrations in the liver while minimizing systemic exposure. This targeted delivery reduces direct exposure to the kidneys. TAF has a longer half-life in the body, meaning it circulates for a longer time but does not require high doses to be effective, thus reducing potential nephrotoxicity. Due to the high efficacy of TAF, a lower dose is typically used, which also helps to reduce potential kidney damage [19]. This is especially critical in patients with decompensated cirrhosis, who often require diuretics for ascites management and are at risk of hepatorenal syndrome. In our study, renal parameters remained stable with TAF in patients with HBV-related decompensated cirrhosis, as with many previous studies. Although there was a slight increase in creatinine in both the TAF and TDF groups after 48 weeks of treatment, it did not yield statistical significance. Regarding the changes in eGFR, a decline was noted in the TDF group (P = 0.343), whereas no significant alterations were detected in the TAF group. These results reinforced TAF’s favorable renal safety profile in patients with decompensated cirrhosis.

While TAF demonstrates favorable renal safety profiles in our cohort, established pharmacovigilance data indicate characteristic adverse effects associated with this agent. TAF-specific safety concerns primarily involve metabolic alterations, including dose-dependent increases in serum lipids (low-density lipoprotein cholesterol and triglycerides) and body weight. These effects are mechanistically linked to TAF’s preferential hepatic metabolism, which reduces systemic tenofovir exposure but may influence lipid regulatory pathways. Additionally, pharmacoepidemiologic studies suggest potential associations between long-term TAF use and insulin resistance perturbations, particularly in patients with preexisting metabolic risk factors [20].

Nevertheless, several limitations must be acknowledged. The study database lacked detailed records on diuretic type or dose, ascites control status, baseline CKD stage, concomitant nephrotoxic agents (e.g., NSAIDs, aminoglycosides, radiographic contrast), and serum albumin—a key marker of volume status and nutritional state. All of these variables are independently associated with TDF- or TAF-related nephrotoxicity and could bias our effect estimates. The study did not collect bone-mineral density data, and the follow-up duration was short (48 weeks); both factors are explicit limitations. The absence of detailed comorbidity data (e.g., metabolic syndromes) may restrict full adjustment for fibrosis determinants, though our propensity matching accounted for all available clinical covariates. Additionally, the lack of a head-to-head randomized controlled trial design precludes definitive conclusions on the superiority of one treatment over the other. Last, the observed effect size may be over-estimated owing to the modest sample, and the true effect of the primary endpoint should be confirmed in prospective, adequately powered trials. Consequently, our results should be interpreted as preliminary signals, and future prospective studies that systematically collect these covariates and apply appropriate adjustment methods are warranted to confirm the advantage of TAF in patients with decompensated cirrhosis.

Conclusion

In conclusion, our study suggests that TAF may offer similar efficacy to TDF in terms of virologic response, with a potentially more favorable safety profile, particularly regarding renal function. Although several comparisons did not reach conventional statistical significance—most likely because of the modest sample size, the retrospective design, and incomplete capture of key nephrotoxic cofactors—the direction and magnitude of the observed effects were consistent across multiple sensitivity analyses and aligned with prior pharmacokinetic and real-world evidence. These converging signals support the contention that, in this high-risk population characterised by diminished renal reserve and heightened susceptibility to drug-related tubular injury, TAF offers a favourable benefit–risk balance. Consequently, we believe that TAF can be reasonably recommended as a first-line antiviral option for treatment-naïve or treatment-experienced patients with decompensated cirrhosis, particularly when renal preservation is a therapeutic priority.

Abbreviations

TAF

Tenofovir alafenamide

TDF

Tenofovir disoproxil fumarate

HBV

Hepatitis B virus

PSM

Propensity score matching

HCC

Hepatocellular carcinoma

ETV

Entecavir

VR

Virologic response

ALT

Alanine Aminotransferase

eGFR

Estimated glomerular filtration rates

APRI

AST to Platelet Ratio Index scores

CHB

Chronic hepatitis B

Authors’ contributions

Lu Chen: Conceived the study, designed the experimental protocols, collected and analyzed the data, and drafted the manuscript.Huifang Zhang: Participated in data collection, and conducted preliminary analysis of the results.Yi-Jing Cai and Lei Zhang: Organized and analyzed the data, and provided data support for manuscript writing.Chao Cai: Conducted literature reviews to provide theoretical background support for the study and proofread and revised the manuscript.Ming-Qin Lu: Was responsible for the ethical review of the study and the handling of related procedures to ensure research compliance and conducted the final review of the manuscript.Yue-Ying Zhu: As the project leader, guided and supervised the entire research process and reviewed and revised the structure and content of the manuscript.

Funding

This work was supported by Key Research and Development Project of Zhejiang Province (No. 2023C03046).

Data availability

The datasets are not publicly available due to privacy regulations. De-identified data can be provided upon reasonable request to the corresponding author (Yue-Ying Zhu, zyy1202@126.com) with approval from the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University.

Declarations

Ethics approval and consent to participate

All procedures were performed in accordance with the guidelines of the institutional ethics committee and adhered to the tenets of the Declaration of Helsinki. This study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University, China, with a waiver of informed consent granted(approval number: KY2024-101). The registration date is April 24, 2024.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.