This study explored differences between ETV and TDF regarding tumor recurrence and death after LT for HBV-related HCC. Patients receiving either NA after LT exhibited similar risks of HCC recurrence, overall death, and HBV recurrence both before and after IPTW. Various Cox models and sensitivity and subgroup analyses confirmed that ETV and TDF were non-inferior to each other with respect to HCC outcomes after LT in HBV-infected patients. Our results provide important insight into the preventive effectiveness of ETV and TDF for HCC, which has been the subject of conflicting reports in the literature.
TDF is often preferred for chronic HBV hepatitis as a naïve treatment and as secondary therapy in patients with a history of drug resistance; however, it has several well-known side effects, including renal tubular dysfunction and decreased bone density (20). International guidelines for treating chronic HBV recommend ETV, instead of TDF, for patients with chronic kidney disease or bone disease, although the efficacy of the two agents is equal when used as primary treatment (21, 22). LT recipients have an increased risk of renal dysfunction and osteoporosis because of their long-term use of tacrolimus and glucocorticoids (23). Recent studies using KOTRY data reported that TDF was associated with a significantly greater decline in estimated glomerular filtration rate than ETV in LT recipients (15). Thus, when HCC outcomes do not differ, ETV could be clinically advantageous over TDF in LT recipients because of its more favorable side effect profile.
The HCC preventive effects of NAs are widely accepted and have been attributed to suppression of the HBV viral load, leading to a reduced liver inflammatory response and less fibrosis (24). In patients with chronic HBV infection, ETV and TDF have similar antiviral effects and likelihood of death or need for LT (22). However, the reported effects of these drugs on the development of HCC have varied between studies. Researchers asserting that TDF is superior to ETV hypothesize that TDF not only suppresses HBV viral replication but also directly suppresses tumor development (11). However, this hypothesis is based on evidence of increased interferon λ-3 with TDF, which has only been shown in experimental models, not in human studies (13). The design of the current study was unique in that it compared the suppressive effects of ETV and TDF on HCC itself (as the patients were HBsAg-negative and had a healthy [transplanted] liver), and the similar risk of HCC recurrence with the two drugs suggests that TDF may not have a direct antitumor effect. However, the median time to HCC recurrence was approximately 12 months after LT, so exposure to NAs was much shorter than the duration of exposure in studies of patients with chronic HBV, necessitating caution when interpreting our results.
In consideration of the substantial discrepancies between previous studies regarding the comparative effects of ETV and TDF on HCC prevention, we applied diverse analytic methods to reduce the possibility of statistical error in the current study. We minimized selection bias between the ETV and TDF groups by using IPTW, and we adjusted for the effects of death from causes other than HCC by performing competing risk regression. Furthermore, we accounted for the effects of changes in immunosuppressant regimens, liver function, and renal function using time-varying Cox models and explored differences between ETV and TDF according to exposure time by performing sensitivity analyses excluding patients who experienced HCC recurrence within 3, 6, or 12 months of LT. The most outstanding aspect of the study was our comparison of ETV and TDF according to the exact HCC burden, which was based on rigorous data collection of explant pathology results. We affirmed that ETV and TDF were mutually non-inferior for HCC outcomes in subgroup analyses stratified by HCC tumor burden.
Tenofovir alafenamide (TAF) is increasingly used for patients with HBV infection because of its lower risk of side effects and similar efficacy, compared with TDF (25). In recent studies of LT recipients, renal function was superior with TAF than with TDF (26). Besifovir is another new NA used to treat HBV, which also has a lower risk of adverse events (27). Although data comparing these new NAs and ETV are currently limited, these new agents are promising treatments for patients with HBV infection who are at considerable risk of renal deterioration or bone disease. No studies have explored the role of TAF or besifovir after LT in patients with HBV-related HCC, and future studies are warranted to compare these two drugs with ETV or TDF for HCC prevention in this patient population.
Many centers use HBIG in addition to high genetic barrier NAs, although the optimal dose and duration of HBIG have not yet been determined. Several studies showed that 5 to 7 days of HBIG plus long-term NAs were sufficient to prevent HBV recurrence (28). A group from Hong Kong even showed reasonable results with HBIG-free HBV prophylaxis after LT, resulting in a 5-year HBV-free survival rate of 87% (29). In Korea, most institutions have been using HBIG after LT for patients with HBV because national insurance covers HBIG daily from the anhepatic phase to 1-week post-LT and weekly thereafter to 1-month, at a dose of 20,000 IU per administration for high-risk patients (HBV DNA-positive or HBeAg-positive) or 10,000 IU per administration for low-risk patients; afterward, treatment is individualized, administered at 1-month or longer intervals (30). HBIG monotherapy without NA is not recommended in Western guidelines because of its relatively low cost-effectiveness (21), but it is the treatment for over 30% of patients in Korea for various reasons, such as insurance problems for using NA in LT recipients who were not on NA before LT (30). Use of HBIG may have contributed to the 5-year HBV-free survival of up to 95% in both the ETV and TDF groups in the current study. The effects of HBIG on HCC outcomes after LT require further study.
The lack of data regarding the exact viral load before LT is a limitation of this study, although it likely had minimum effect on our results because most patients were persistently HBsAg-negative, probably because of the use of HBIG. Another limitation was a lack of information about the use of NAs before and after LT. Insurance coverage is the most important factor for selecting HBV prophylaxis drugs in Korea, and only the same NA used before LT is currently covered by national insurance. Thus, patients who were receiving ETV or TDF because of renal function, bone density, history of drug resistance, or physician preference before LT might be maintained on the same drug after LT in our study population. Data regarding this issue were not available in the KOTRY database, but it is unlikely that these data would have affected the post-LT results. Lastly, our results were based on a predominantly living donor LT population, so caution is required in generalizing our results, especially in Western countries where mainstream management of patients with HCC involves down-staging and waiting for deceased donor LT.
Despite these limitations, to our knowledge, this study is the first to compare the effects of ETV versus TDF on HCC recurrence after LT using large multicenter data. ETV and TDF showed mutual non-inferiority for HCC outcomes when used for HBV prophylaxis after LT.