Opinion

EDITORIAL

Tenofovir vs Entecavir for Hepatocellular Carcinoma Prevention in Patients With Chronic B One of These Things Is Not Like the Other Jennifer A. Flemming, MD, MAS; Norah A. Terrault, MD, MPH

Current guidelines for the treatment of chronic of propensity scores and competing risks, as well as validation in (CHB) recommend as first-line line therapy entecavir (ETV), a hospital-based cohort having data on disease severity and vi- fumarate (TDF), or rologic responses. Remarkably,the association of ETV vs TDF and fumarate.1-3 Both ETV and TDF HCC risk was minimally changed by any of these adjustments. achieve similar rates of hepa- However, adherence to antiviral therapy and surveillance pro- Related article page 30 titis B virus (HBV) DNA sup- grams was not captured and may have influenced rates of HCC. pression and alanine amino- Indeed, adherence of less than 90% to antiviral therapy in patients transferase normalization over time and have an excellent safety with CHB has been associated with a higher risk of HCC.10 More- record. However, ETV and TDF have not been directly compared over, especially in countries where prior exposure is except in patients with decompensated cirrhosis4; to our knowl- prevalent, patients treated with ETV may be at increased risk of edge, no other head-to-head randomized trials exist. Yet, there virologic breakthrough with time, and this, in turn, affects risk is strong interest in understanding if one drug might be supe- for HCC. Indeed, in the study by Choi and colleagues, there was rior to the other in specific clinical settings or subpopulations. a substantially higher rate of changing therapy in patients treated In the absence of randomized comparator trials, high-quality with ETV (12%) vs TDF (0.2%). Might reduced adherence and in- “real world” cohorts can be quite informative. In this issue of creasing rates of ETV resistance represent unmeasured confound- JAMA Oncology, Choi and colleagues5 report the results of 2 ret- ers that contribute to the divergence in rates of HCC evident 2 years rospective cohort studies of patients with CHB from Korea that after initiation of therapy? Ideally, a randomized study design evaluated the association between the type of nucleoside or would help overcome these challenges and provide stronger evi- analogue and the development of hepatocellular car- dence of the association between type of antiviral agent and HCC cinoma (HCC).5 Using both a large administrative data set (24 156 outcomes. As such a study seems unlikely,at least in the near term, patients with CHB and 984 with HCC) and a tertiary hospital- we would advocate for additional observational cohorts with data based cohort (2701 patients with CHB and 154 with HCC), the on liver disease severity, virologic response, and adherence to authors found that the risk of HCC was consistently about 35% surveillance to evaluate this important question. lower in those treated with TDF than in those treated with ETV. In considering HBV therapy as prevention of HCC, a critical This association persisted even when stratified by question remains. How important is the achievement and main- status and after very thorough adjustment for available con- tenance of undetectable HBV DNA during treatment? Current founders. These striking findings raise the question of whether HBV guidelines do not recommend changing or adding another the current CHB guidelines should be updated to reflect this antiviral agent in patients with low-level intermittent or persis- observed superior anticancer benefit of TDF over ETV. tent viremia during treatment with first-line agents, as long as Multiple studies have shown that treatment with either ETV criteria for virologic breakthrough are not met. In terms of HCC or TDF reduces the incidence of HCC in patients with CHB when risk, natural history studies show that HCC risk is highest among compared with untreated patients.6,7 Prior studies examining those with HBV DNA levels of 200 000 IU/mL or higher but the use of ETV vs TDF,though limited in numbers, did not find not different from uninfected controls if HBV DNA levels are lower a difference between these 2 first-line agents.8,9 Based on the than 2000 IU/mL.11 More recent studies of patients undergoing observed adjusted hazard ratios in the study by Choi and long-term antiviral therapy provide conflicting results on the colleagues,5 these other cohort studies would have needed more importance of low-level viremia and HCC risk. In a Korean study than 200 cases of HCC to have adequate power to detect a 35% of patients with CHB and cirrhosis on ETV therapy for up to 5 reduction in HCC risk. Therefore, since the largest known study years, rates of HCC were 2.2-fold higher in patients with persis- had only 56 incident cases of HCC,8 these earlier studies may tent or intermittent HBV DNA viral loads less than 2000 IU/mL have been underpowered to detect a difference in HCC rates compared with those with persistently undetectable HBV DNA by . (<12 IU/mL), though this did not hold true for those without There remains a concern that the observed difference in HCC cirrhosis.12 In a study of patients with decompensated cirrhosis, rates is due to residual confounding. This is a limitation of all ob- maintained virologic response (<20 IU/mL) vs suboptimal re- servational studies, but Choi and colleagues5 used a robust ana- sponses (<2000 but >20 IU/mL) influenced transplant-free lytic approach that included multiple subgroup analyses, the use survival but not HCC risk over 10 years.13 These studies highlight

jamaoncology.com (Reprinted) JAMA Oncology January 2019 Volume 5, Number 1 17

© 2018 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Jules Levin on 04/15/2019 Opinion Editorial

the importance of cirrhosis as a primary driver of HCC risk, visual inspection of the survival curves in ETV-treated patients independent of effective therapy.12,13 However, for those with- suggests that the rate of incident HCC is not increasing. Another out cirrhosis, which represent the majority of patients with Korean cohort of almost 2000 patients treated with ETV showed CHB, the level of viremia may be relevant to treatment decisions. that the annual HCC incidence in the first 5 years of treatment While the study by Choi et al5 includes information on virologic was similar to the rate after 5 years of ETV exposure (2.29% vs response in the first year with the validation cohort, the absence 1.66%, P = .22).14 Additionally, in a cohort of over 1900 white of longitudinal virologic response over the duration of follow-up patients with CHB undergoing long-term treatment with either is a shortcoming of the larger observational cohort, and could be ETV or TDF with a median follow-up of 6 years, HCC incidence the major explanatory variable for the differences in HCC out- was similar beyond 5 years in those without cirrhosis, while comes in the ETV vs TDF groups. Collectively,recent studies sug- there was a significant decrease in HCC incidence after 5 years gest that viral suppression is important, and thus selecting an of therapy in those with cirrhosis at baseline.6 Thus, there is no antiviral agent that achieves and maintains suppression is highly clinical support for a procarcinogenic tendency with ETV,and desirable, but the findings do not dictate that one drug should the biologic underpinnings of the observed difference in HCC be preferred over the other. rates between ETV and TDF remain obscure. Finally, if TDF is superior to ETV in reducing HCC risk, the Three safe and effective oral antivirals are endorsed by ex- big question is why. Choi and colleagues5 offer 2 potential ex- perts as first-line therapy for CHB: ETV,TDF,and tenofovir ala- planations, neither of which is well supported by human data. fenamide fumarate.1,2 These antiviral agents share the quali- They suggest that ETV may be procarcinogenic and/or that ties of high efficacy in reducing HBV DNA levels and excellent nucleotide analogues (vs nucleoside analogues) may induce tolerability and safety. In deciding on the best option for a given higher antitumor cytokines. That viral clearance may induce patient with CHB, the clinician should consider the treatment an altered intrahepatic milieu that influences risk of hepatocar- history, comorbidities, including renal disease or human im- cinogenesis is possible, though the rapidity of viral and antigen munodeficiency virus, and cost and accessibility. Prevention of decline may be more important than the specific drug type liver-related complications requires adherence to the treat- (nucleoside vs nucleotide) used. The hypothesis that ETV may ment regimen and a therapy achieving viral load suppression. be procarcinogenic is based on preclinical studies in mice and The work of Choi and colleagues5 is the first to our knowledge rats treated with high-dose ETV; there are no human correla- to suggest that TDF offers advantages over ETV in terms of HCC tive data to support this. Indeed, if this were the case, one would prevention, but given the inherent limitations of observa- have expected to see increasing rates of HCC with increasing tional data, this study should not lead to a widespread para- duration of drug exposure; a feature not seen in this cohort or digm shift in selecting TDF over ETV.Rather, it should spur more others.5,6,14 The study by Choi et al with just over 4 years of investigators to address the issue of whether one of these HBV follow-up data does not provide a yearly HCC incidence, but drugs is not like the other.

ARTICLE INFORMATION hepatitis B: a 2015 update. Hepatol Int. 2016;10(1): virus-related cirrhosis. Indian J Gastroenterol. 2015; AuthorAffiliations:DepartmentofMedicine,Queen’s 1-98. doi:10.1007/s12072-015-9675-4 34(4):286-291. doi:10.1007/s12664-015-0576-1 University, Kingston, Ontario, Canada (Flemming); 4. Liaw YF, Sheen IS, Lee CM, et al. Tenofovir 10. Shin JW, Jung SW, Lee SB, et al. Medication Department of Public Health Sciences, Queen’s disoproxil fumarate (TDF), emtricitabine/TDF, and nonadherence increases hepatocellular carcinoma, University, Kingston, Ontario, Canada (Flemming); entecavir in patients with decompensated chronic cirrhotic complications, and mortality in chronic Department of Medicine, University of California, hepatitis B liver disease. Hepatology. 2011;53(1): hepatitis B patients treated with entecavir. Am J San Francisco (Terrault); Department of Transplant 62-72. doi:10.1002/hep.23952 Gastroenterol. 2018;113(7):998-1008. doi:10.1038/ Surgery, University of California, San Francisco 5. Choi J, Kim HJ, Lee J, et al. Risk of hepatocellular s41395-018-0093-9 (Terrault). carcinomainpatientstreatedwithentecavirvstenofovir 11. Chen CJ, Yang HI, Su J, et al; REVEAL-HBV Study Corresponding Author: Norah A. Terrault, MD, forchronichepatitisB:aKoreannationwidecohortstudy Group. Risk of hepatocellular carcinoma across a MPH, University of California, San Francisco, 513 [publishedonlineSeptember27,2018].JAMAOncol.doi: biological gradient of serum DNA Parnassus Ave, Room S357, San Francisco, CA 94143 10.1001/jamaoncol.2018.4070 level. JAMA. 2006;295(1):65-73. doi:10.1001/jama. ([email protected]). 6. Papatheodoridis GV, Idilman R, Dalekos GN, et al. 295.1.65 Published Online: September 27, 2018. The risk of hepatocellular carcinoma decreases after 12. Kim JH, Sinn DH, Kang W, et al. Low-level doi:10.1001/jamaoncol.2018.4039 the first 5 years of entecavir or tenofovir in Caucasians viremia and the increased risk of hepatocellular Conflict of Interest Disclosures: None reported. with chronic hepatitis B. Hepatology. 2017;66(5): carcinoma in patients receiving entecavir 1444-1453. doi:10.1002/hep.29320 treatment. Hepatology. 2017;66(2):335-343. doi: 10.1002/hep.28916 REFERENCES 7. Kim BG, Park NH, Lee SB, et al. Mortality, liver transplantation, and hepatic complications in patients 13. Jang JW, Choi JY, Kim YS, et al. Effects of 1. Terrault NA, Lok ASF, McMahon BJ, et al. Update on with treatment-naïve chronic hepatitis B treated with virologic response to treatment on short- and prevention, diagnosis, and treatment of chronic hepa- entecavir vs tenofovir [published online July 11, 2018]. long-term outcomes of patients with chronic titis B: AASLD 2018 hepatitis B guidance. Hepatology. J Viral Hepat. 2018. doi:10.1111/jvh.12971 hepatitis B virus infection and decompensated 2018;67(4):1560-1599. doi:10.1002/hep.29800 8. Papatheodoridis GV, Dalekos GN, Yurdaydin C, cirrhosis. Clin Gastroenterol Hepatol. 2018; 2. European Association for the Study of the Liver. et al. Incidence and predictors of hepatocellular S1542-3565(18)30471-3. EASL clinical practice guidelines: management of carcinoma in Caucasian chronic hepatitis B patients 14. Kim BG, Park NH, Lee SB, et al. The risk of hepato- chronic hepatitis B virus infection. J Hepatol. 2012; receiving entecavir or tenofovir. J Hepatol. 2015; cellular carcinoma within and beyond the first 5 years 57(1):167-185. doi:10.1016/j.jhep.2012.02.010 62(2):363-370. doi:10.1016/j.jhep.2014.08.045 of entecavir in Korean patients with chronic hepatitis . Sarin SK, Kumar M, Lau GK, et al. Asian-Pacific 3 9. Goyal SK, Dixit VK, Shukla SK, et al. Prolonged B [published online July 27,2018]. Liver Int. 2018. doi: clinical practice guidelines on the management of use of tenofovir and entecavir in hepatitis B 10.1111/liv.13938

18 JAMA Oncology January 2019 Volume 5, Number 1 (Reprinted) jamaoncology.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Jules Levin on 04/15/2019