Letters to the Editors
Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? M. D. Sadler & S. S. Lee University of Calgary Liver Unit, Calgary, AB, Canada. E-mail: [email protected] doi:10.1111/apt.12732
SIRS, Gao et al. concluded that tenofovir (TDF) was superior to entecavir (ETV) in treating HBeAg-positive chronic hepatitis B infection (CHB).1 However, the conclusions made in this study may be misleading due to important limitations in how these drugs were compared. This was not a prospective randomised head-to-head assessment of the efficacy of these two drugs, but rather a retrospective analysis of nonrandomised groups of patients. Thus, hidden biases cannot be ruled out. Making comparisons of treatment efficacy in this way may result in faulty conclusions, which lack the power to clearly guide treatment decisions. Although potentially a reflection of real-world experience, we were surprised to see such low response rates for these drugs. Randomised studies of ETV and TDF have shown undetectable HBV DNA levels after 48 weeks of therapy in HBeAg+ patients in 67%2 and 76%3 of patients, respectively, whereas Gao et al. report response rates as low 28% for ETV and 51% for TDF after 12 months of treatment. Moreover, the randomised trials of both ETV and TDF primarily assessed responses in patients with high baseline viral loads.2, 3
Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? Authors’ reply L. Gao*, H. N. Trinh†, J. Li‡ & M. H. Nguyen§ *Department of Chemistry, Stanford University, Stanford, CA, USA. † San Jose Gastroenterology, San Jose, CA, USA. ‡ Palo Alto Medical Foundation, Mountain View, CA, USA. § Division of Gastroenterology and Hepatology, Stanford University Medical Center, Stanford, CA, USA. E-mail: [email protected] doi:10.1111/apt.12765
Aliment Pharmacol Ther 2014; 39: 1337-1344 ª 2014 John Wiley & Sons Ltd
The conclusions are based on a very small subset of patients. Only 13 patients in the entire HBeAg+ TDF group and 6 patients in the HBeAg+ TDF group with baseline HBV DNA greater than 8 log10 IU/mL were followed up long enough to be included in the analysis after 12 months of therapy. Furthermore, the lower limit of time for follow-up in either ETV- or TDF-treated HBeAg+ patients was less than 6 months. Too few patients were followed up for too short a duration to allow useful interpretations of these results. As a result of all of this, the results of this study need to be validated in a direct randomised comparison of ETV and TDF, before they can be applied in the management of patients with CHB.
ACKNOWLEDGEMENTS Declaration of personal interests: SSL has received research funding, consultation fees and speaker’s honoraria from Bristol Myers Squibb and Gilead. Declaration of funding interests: None. REFERENCES 1. Gao L, Trinh HN, Li J, Nguyen MH. Tenofovir is superior to entecavir for achieving complete viral suppression in HBeAgpositive chronic hepatitis B patients with high HBV DNA. Aliment Pharmacol Ther 2014; 39: 629–37. 2. Chang T-T, Gish RG, de Man R, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006; 354: 1001–10. 3. Marcellin P, Heathcote EJ, Buti M, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med 2008; 359: 2442–55.
SIRS, We respect the comments by Lee and Sadler and welcome discussion and critical review of our results.1, 2 We were aware of both the clinical trials3, 4 referenced in the commentary, and we cited and discussed them in the original paper.2 There were a few notable differences between our study and the clinical trials cited,3, 4 which may partly explain the discrepancy in results, as we have discussed.2 First, these studies3, 4 and our study both indeed had patient populations with high viral loads on average. However, unlike in the clinical trials,3, 4 we excluded those with lower baseline viral loads. Our results are consistent with Gordon et al.,5 who reported approximately
1339
Letters to the Editors 42% complete viral suppression at 48 weeks for tenofovir patients when restricted to those with baseline viral loads >9 log10 copies/mL. Second, our study focused on a real-life treatment setting; this is in contrast to a clinical trial, in which treatment protocols are followed much more rigorously and higher response rates are expected. Finally, we agree that the number of patients in Figure 2 was low.2 However, this was only used as a supplement to the main results (Figure 1 and Table 3), which were based on greater numbers of patients.
ACKNOWLEDGEMENT The authors’ declarations of personal and financial interests are unchanged from those in the original article.2
Letter: prognostic scores in alcoholic hepatitis E. Forrest Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, UK. E-mail: [email protected] doi:10.1111/apt.12734
SIRS, The paper by Papastergiou et al. assesses the utility of the proliferating number of clinical scores used to determine prognosis in alcoholic hepatitis.1 The difference between this analysis and other similar studies is that it uses a biopsy-proven cohort of patients. Despite this, it comes to the same conclusion as other nonbiopsy studies, namely that there is no clear difference between the scores in determining outcome.2 However to interpret this result it is important to remember the reasons for developing a clinically useful score in the first place: it is not simply a case of identifying patients who have a poor prognosis. A clinically relevant score should also identify those patients who will benefit most from a therapeutic intervention. In keeping with most other studies that have assessed these scores, the authors have analysed patients, 48% of whom received disease modifying treatment in the form of corticosteroids. If a score is effective in identifying those whose outcome improves with such treatment then inevitably the prognostic utility of that score will be reduced if treated patients are included in the analysis. The discriminant factor (DF) and the Glasgow alcoholic hepatitis score 1340
REFERENCES 1. Sadler MD, Lee SS. Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? Aliment Pharmacol Ther 2014; 39: 1339. 2. Gao L, Trinh HN, Li J, Nguyen MH. Tenofovir is superior to entecavir for achieving complete viral suppression in HBeAgpositive chronic hepatitis B patients with high HBV DNA. Aliment Pharmacol Ther 2014; 39: 629–37. 3. Chang TT, Gish RG, de Man RA, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006; 354: 1001–10. 4. Marcellin P, Heathcote EJ, Buti M, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med 2008; 359: 2442–55. 5. Gordon SC, Krastev Z, Horban A, et al. Efficacy of tenofovir disoproxil fumarate at 240 weeks in patients with chronic hepatitis B with high baseline viral load. Hepatology 2013; 58: 505–13.
(GAHS) differ from the other scores described in this regard as they have been either derived from, or tested upon, untreated patients. The DF appears to indicate patients who benefit from corticosteroid therapy in clinical trials.3 The GAHS was derived from a large untreated population of patients,4 and has subsequently been shown to identify those who benefit from treatment with greater specificity than the DF.5, 6 A study of biopsy-proven patients may be thought to be more accurate. However, there are also dangers in this approach, as biopsy appearances are not specific to alcoholic hepatitis. In a study of alcohol-related acute-on-chronic liver failure a biopsied control group of patients with chronic decompensation, in whom continuing alcohol abuse was ruled out, was found to have features of ‘alcoholic steatohepatitis’ in 94%: 27% described as moderate or severe.7 This included 78% with hepatocyte ballooning and 67% with megamitochondria. Thus, patients with ‘biopsy-proven’ alcoholic hepatitis may in fact not have the condition in any clinical sense. This may be the case for some of the patients in this study as Table 2 indicated that some patients had serum bilirubin values as low as 44lmol/L, well below the threshold for inclusion in recent clinical trials of alcoholic hepatitis.8, 9 Thus, while reassuring us that the available prognostic scores have excellent negative predictive values, the testing of such scores on a treated population may underestimate the utility of those scores which can direct clinical treatment, and the use of a biopsy-proven cohort might still lead to the inclusion of patients without clinical alcoholic hepatitis. Aliment Pharmacol Ther 2014; 39: 1337-1344 ª 2014 John Wiley & Sons Ltd
Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? M. D. Sadler & S. S. Lee University of Calgary Liver Unit, Calgary, AB, Canada. E-mail: [email protected] doi:10.1111/apt.12732
SIRS, Gao et al. concluded that tenofovir (TDF) was superior to entecavir (ETV) in treating HBeAg-positive chronic hepatitis B infection (CHB).1 However, the conclusions made in this study may be misleading due to important limitations in how these drugs were compared. This was not a prospective randomised head-to-head assessment of the efficacy of these two drugs, but rather a retrospective analysis of nonrandomised groups of patients. Thus, hidden biases cannot be ruled out. Making comparisons of treatment efficacy in this way may result in faulty conclusions, which lack the power to clearly guide treatment decisions. Although potentially a reflection of real-world experience, we were surprised to see such low response rates for these drugs. Randomised studies of ETV and TDF have shown undetectable HBV DNA levels after 48 weeks of therapy in HBeAg+ patients in 67%2 and 76%3 of patients, respectively, whereas Gao et al. report response rates as low 28% for ETV and 51% for TDF after 12 months of treatment. Moreover, the randomised trials of both ETV and TDF primarily assessed responses in patients with high baseline viral loads.2, 3
Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? Authors’ reply L. Gao*, H. N. Trinh†, J. Li‡ & M. H. Nguyen§ *Department of Chemistry, Stanford University, Stanford, CA, USA. † San Jose Gastroenterology, San Jose, CA, USA. ‡ Palo Alto Medical Foundation, Mountain View, CA, USA. § Division of Gastroenterology and Hepatology, Stanford University Medical Center, Stanford, CA, USA. E-mail: [email protected] doi:10.1111/apt.12765
Aliment Pharmacol Ther 2014; 39: 1337-1344 ª 2014 John Wiley & Sons Ltd
The conclusions are based on a very small subset of patients. Only 13 patients in the entire HBeAg+ TDF group and 6 patients in the HBeAg+ TDF group with baseline HBV DNA greater than 8 log10 IU/mL were followed up long enough to be included in the analysis after 12 months of therapy. Furthermore, the lower limit of time for follow-up in either ETV- or TDF-treated HBeAg+ patients was less than 6 months. Too few patients were followed up for too short a duration to allow useful interpretations of these results. As a result of all of this, the results of this study need to be validated in a direct randomised comparison of ETV and TDF, before they can be applied in the management of patients with CHB.
ACKNOWLEDGEMENTS Declaration of personal interests: SSL has received research funding, consultation fees and speaker’s honoraria from Bristol Myers Squibb and Gilead. Declaration of funding interests: None. REFERENCES 1. Gao L, Trinh HN, Li J, Nguyen MH. Tenofovir is superior to entecavir for achieving complete viral suppression in HBeAgpositive chronic hepatitis B patients with high HBV DNA. Aliment Pharmacol Ther 2014; 39: 629–37. 2. Chang T-T, Gish RG, de Man R, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006; 354: 1001–10. 3. Marcellin P, Heathcote EJ, Buti M, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med 2008; 359: 2442–55.
SIRS, We respect the comments by Lee and Sadler and welcome discussion and critical review of our results.1, 2 We were aware of both the clinical trials3, 4 referenced in the commentary, and we cited and discussed them in the original paper.2 There were a few notable differences between our study and the clinical trials cited,3, 4 which may partly explain the discrepancy in results, as we have discussed.2 First, these studies3, 4 and our study both indeed had patient populations with high viral loads on average. However, unlike in the clinical trials,3, 4 we excluded those with lower baseline viral loads. Our results are consistent with Gordon et al.,5 who reported approximately
1339
Letters to the Editors 42% complete viral suppression at 48 weeks for tenofovir patients when restricted to those with baseline viral loads >9 log10 copies/mL. Second, our study focused on a real-life treatment setting; this is in contrast to a clinical trial, in which treatment protocols are followed much more rigorously and higher response rates are expected. Finally, we agree that the number of patients in Figure 2 was low.2 However, this was only used as a supplement to the main results (Figure 1 and Table 3), which were based on greater numbers of patients.
ACKNOWLEDGEMENT The authors’ declarations of personal and financial interests are unchanged from those in the original article.2
Letter: prognostic scores in alcoholic hepatitis E. Forrest Department of Gastroenterology, Glasgow Royal Infirmary, Glasgow, UK. E-mail: [email protected] doi:10.1111/apt.12734
SIRS, The paper by Papastergiou et al. assesses the utility of the proliferating number of clinical scores used to determine prognosis in alcoholic hepatitis.1 The difference between this analysis and other similar studies is that it uses a biopsy-proven cohort of patients. Despite this, it comes to the same conclusion as other nonbiopsy studies, namely that there is no clear difference between the scores in determining outcome.2 However to interpret this result it is important to remember the reasons for developing a clinically useful score in the first place: it is not simply a case of identifying patients who have a poor prognosis. A clinically relevant score should also identify those patients who will benefit most from a therapeutic intervention. In keeping with most other studies that have assessed these scores, the authors have analysed patients, 48% of whom received disease modifying treatment in the form of corticosteroids. If a score is effective in identifying those whose outcome improves with such treatment then inevitably the prognostic utility of that score will be reduced if treated patients are included in the analysis. The discriminant factor (DF) and the Glasgow alcoholic hepatitis score 1340
REFERENCES 1. Sadler MD, Lee SS. Letter: treatment of HBeAg+ chronic hepatitis B – is tenofovir truly superior to entecavir? Aliment Pharmacol Ther 2014; 39: 1339. 2. Gao L, Trinh HN, Li J, Nguyen MH. Tenofovir is superior to entecavir for achieving complete viral suppression in HBeAgpositive chronic hepatitis B patients with high HBV DNA. Aliment Pharmacol Ther 2014; 39: 629–37. 3. Chang TT, Gish RG, de Man RA, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006; 354: 1001–10. 4. Marcellin P, Heathcote EJ, Buti M, et al. Tenofovir disoproxil fumarate versus adefovir dipivoxil for chronic hepatitis B. N Engl J Med 2008; 359: 2442–55. 5. Gordon SC, Krastev Z, Horban A, et al. Efficacy of tenofovir disoproxil fumarate at 240 weeks in patients with chronic hepatitis B with high baseline viral load. Hepatology 2013; 58: 505–13.
(GAHS) differ from the other scores described in this regard as they have been either derived from, or tested upon, untreated patients. The DF appears to indicate patients who benefit from corticosteroid therapy in clinical trials.3 The GAHS was derived from a large untreated population of patients,4 and has subsequently been shown to identify those who benefit from treatment with greater specificity than the DF.5, 6 A study of biopsy-proven patients may be thought to be more accurate. However, there are also dangers in this approach, as biopsy appearances are not specific to alcoholic hepatitis. In a study of alcohol-related acute-on-chronic liver failure a biopsied control group of patients with chronic decompensation, in whom continuing alcohol abuse was ruled out, was found to have features of ‘alcoholic steatohepatitis’ in 94%: 27% described as moderate or severe.7 This included 78% with hepatocyte ballooning and 67% with megamitochondria. Thus, patients with ‘biopsy-proven’ alcoholic hepatitis may in fact not have the condition in any clinical sense. This may be the case for some of the patients in this study as Table 2 indicated that some patients had serum bilirubin values as low as 44lmol/L, well below the threshold for inclusion in recent clinical trials of alcoholic hepatitis.8, 9 Thus, while reassuring us that the available prognostic scores have excellent negative predictive values, the testing of such scores on a treated population may underestimate the utility of those scores which can direct clinical treatment, and the use of a biopsy-proven cohort might still lead to the inclusion of patients without clinical alcoholic hepatitis. Aliment Pharmacol Ther 2014; 39: 1337-1344 ª 2014 John Wiley & Sons Ltd
