- Email: [email protected]
Treatment of chronic hepatitis B in children with prednisone followed by alfa-interferon: a controlled randomized study
Journal o f Hepatology 1994; 20:163-167 Printed in Denmark, All rights reserved Munksgaard. Copenhagen
Copyright © Journalof Hepatology 1994
Journal of Hepatology ISSN 0168-8278
Treatment of chronic hepatitis B in children with prednisone followed by alfa-interferon: a controlled randomized study R i c c a r d o Utili t, Evangelista Sagnelli 2, G i o v a n n i B. G a e t a 2, B r u n o Galanti 2, Salvatore Nardiello 2, Francesca M. Felaco 2, Giuseppe Pasquale 2, Aldo M a r r o n e ~, Lucia A p r e a 2, Teresa Pizzella 2, L a u r a Digilio% Giuseppe Cesaro', Carolina S a r d a r o 2, Lucio Santarpia ~, Luigi E. Adinolfi', Augusto Andreana% Michele Russo 2, Pietro Filippini 2, R o s a Z a m p i n o l, Felice Piccinino 3, Giuseppe Ruggiero t and Guiseppe Giusti 2 Ilnstitute of Medical Therapy, :Institute of lnfectious Diseases and Jlnstitute of Tropical and Subtropical Diseases, University of Naples, 2nd Medical School, Naples, Italy
(Received 10 April 1992)
The efficacy and safety of sequential treatment with prednisone and interferon was evaluated in a randomized, controlled study on 43 children with biopsy proven HBsAg/HBeAg/hepatitis B virus-DNA positive, anti-delta negative, chronic hepatitis (34 chronic persistent hepatitis, 9 chronic active hepatitis). Patients received either a 1-month course of prednisone (0.6 to 0.3 mg/kg per day) followed by interferon alfa-2a (3 MU/m% thrice weekly, for 12 months; 22 patients) or no treatment (21 patients). At the end of the study (20 months), clearance of hepatitis B virus-DNA and HBeAg seroconversion were observed in nine (41%) of the patients treated with prednisone and interferon and in two (9.5%) of the untreated controls (p=0.020). Two of the treated patients who lost HBeAg, also cleared HBsAg. In the treated group, 13 (59%) patients had stable normal levels of alanine aminotransferase on their last examination. The baseline serum level of hepatitis B virus-DNA was an important predictor of response. In fact, HBeAg clearance was observed in 75% of patients with a baseline hepatitis B virusDNA level lower than 100 pg/ml and in none with a level above 100 pg/ml. We suggest that combined treatment with prednisone followed by alfa-interferon may be safe and effective in inducing a stable clearance of HBeAg and, in some cases, of HBsAg in children with chronic hepatitis B and with a low level of viral replication. For children with high levels of viral replication, this regimen seems to be ineffective. © Journal of Hepatology. Key words: Childhood hepatitis B; Chronic hepatitis B; Interferon; Prednisone withdrawal
Chronic hepatitis B in children is usually an asymptomatic disease, associated in most cases with a high level of hepatitis B virus (HBV) replication and low inflammatory activity (1). Nevertheless, these children may experience a flare-up of the diseases which may induce a rapid progression to a more severe form of chronic liver disease and in some cases hepatocellular carcinoma (2,3). Therefore, antiviral treatment for this condition appears warranted. The aim of treatment with alfa interferon (IFN) in these patients is to reduce viral replication and to induce a moderate activation of the immune response to HBV, which may lead to a self-limiting necrosis of the infected
hepatocytes. Consequently, a stable suppression of viral replication (clearance of HBeAg and HBV-DNA) may be expected and, in a few cases, a complete resolution of HBV infection (clearance of HBsAg) may follow. IFN treatment seems to be effective for adults with chronic hepatitis B (4,5). The same treatment, however, seems to be less effective for children (6,7)_ This has been attributed to an immunological tolerance to the virus which, in most cases, was acquired at an early age when the immune system was immature (8). It has been proposed that the response rate to IFN may be enhanced by a priming course of corticosteroids (9).
Correspondence to: Prof. R. Utili, Istituto di Terapia Medica, I Facoltfi di Medicina, UniversitY.di Napoli, Via Cotugno 1 (c/o Ospedale Ges/J e Maria), 80135 Naples, Italy
164 The rationale for this approach is based on the fact that the withdrawal of corticosteroid therapy may induce an immunological rebound, followed by an episode of liver cell necrosis and by a decrease in viral replication (10). This treatment regimen was shown to have a slight advantage over IFN alone in inducing a state of viral suppression in Chinese children (8). There are no similar studies in Caucasian children. Accordingly, we initiated a randomized controlled study to evaluate the efficacy of sequential treatment with a short course of prednisone followed by a long course of IFN in young patients with chronic hepatitis and active HBV replication.
Patients and Methods
Criteria for inclusion in the study were: age below 16 years, infection by HBV with markers of active viral replication in serum (both HBeAg and HBV-DNA) and persistent elevation of serum aminotransferase values (at least 1.5 times the upper normal limit) for at least 6 months; evidence of chronic hepatitis on liver biopsy performed in the preceding 3 months. Exclusion criteria were: history of acute hepatitis in the previous 24 months, previous antiviral or immunosuppressive therapy, presence of antibodies to hepatitis delta virus (HDV), hepatitis C virus (HCV), or human immunodeficiency virus (HIV), decompensated liver disease, white-cell counts below 3x 109 per liter, platelet counts below 70x 109 per liter, drug addiction, or any other serious illnesses. According to these criteria, 43 Italian patients, 6-16 years old, entered the study between September 1989 and November 1990. They were randomly assigned to two groups according to a simple randomization list prepared in advance from a random digit table. Consecutively numbered sealed envelopes with the treatment to be assigned were used to preserve the unpredictability of the next assignment. The first group consisted of 21 patients who were left untreated and served as controls. The second group consisted of 22 patients who received a lmonth course of prednisone, 0.6 mg/kg per day orally for the first 3 weeks and 0.3 mg/kg per day in the 4th week, followed by a 2-week rest and then, treatment with interferon alfa-2a 3 MU/m 2 three times weekly for 12 months. All patients were examined at 2-week intervals during the month of prednisone treatment and during the first 2 months of interferon treatment. Then they were examined at 1-month intervals during the interferon treatment and for the next 6 months after IFN discontinuation; the total follow-up was at least 20 months. Control patients were also evaluated at monthly intervals. On each examination, the patients received a clinical, biochemical, hematolog-
R. UTILI et al. ical and virological assessment and any adverse reactions or side effects of treatment were registered. Interferon dosage was reduced or discontinued in the event of serious side effects. Informed or written consent was obtained from a parent of each patient. HBsAg, HBsAb, HBeAg, HBeAb, and anti-HDV were tested with commercial enzyme-linked immunosorbent assays (Abbott). HCV-Ab was tested by Ortho-HCV ELISA Test System 2nd generation and HIV-Ab by HIV-1 Recombinant (Abbott). The HBV-DNA was detected by slot-hybridization using a 32P-labelled cloned HBV-DNA probe as previously described (6). Semi-quantitation of HBV-DNA was obtained by comparing the autoradiographic signals of appropriately diluted serum samples to those of serial dilutions of a standard serum with known concentrations of HBV-DNA (Genostics Hepatititis B Virus D N A positive control, Abbott). Test sensitivity was 1 pg of HBV-DNA per ml of serum (11). The outcome of the disease was assessed at the end of the observation period. Response to treatment was defined as follows: 1) full, in the presence of irreversible disappearance of HBV-DNA during therapy associated with a stable HBeAg/HBeAb seroconversion which persisted for at least 6 months after the end of therapy, with or without HBsAg clearance; 2) partial, when a loss or at least a stable 75% decrease in serum HBV-DNA concentration occurred during therapy, with persistence of HBeAg; 3) absent, when no change in serum HBV-DNA levels was observed during treatment. No second liver biopsy was performed at the end of the study for ethical reasons. Statistics The results are expressed as mean---SD or, when indicated, as the median value. Analysis of HBV-DNA levels was made after a square-root transformation to achieve normally distributed data. The percentage differences were analyzed by the Fisher exact test. The independent predictors of response were evaluated by the chi-square test for heterogeneity (12).
Results
All 43 patients were asymptomatic on admission and their disease was generally mild, as demonstrated by the predominance of chronic persistent hepatitis vs chronic active hepatitis (Table l). Fifteen (eight controls, seven treated) were born to HBsAg-positive mothers and, presumably, they had acquired the infection by vertical transmission. The other 28 patients belonged to families with at least one relative, other than the mother, infected by HBV. Five of them (one control and four
PREDNISONE-IFN IN CHILDHOOD HEPATITIS B treated) had a past history o f s y m p t o m a t i c acute viral hepatitis. The control and treated groups were c o m p a r a b l e with respect to age, sex ratio, biochemical and histologic features and baseline H B V - D N A (Table 1). According to the study protocol, a full response (permanent clearance o f H B V - D N A and H B e A g seroconversion) was observed in nine (41%) patients treated with prednisone plus interferon, and in two (9.5%) of the untreated controls (p=0.020 by the Fisher exact test). A m o n g treated patients who responded to treatment, two also cleared HBsAg. All patients who responded to treatment had a peak elevation o f alanine aminotransferase (ALT) activity (from 4 to 25 times upper n o r m a l levels) within the 2rid m o n t h of treatment with interferon, and cleared H B V - D N A shortly thereafter_ In eight responders, seroconversion to H B e A b occurred from 1 to 9 months after the initiation o f I F N treatment. The ninth responder patient seroconverted to H B e A b 2 months after treatment was discontinued. In one patient who had shown H B e A g seroconversion early on in I F N treatment, H B e A g and H B V - D N A reappeared at the end o f I F N administration. However, 2 months later he had an increase in A L T values with a stable clearance of HBV-
TABLE I Characteristics of patients on admission Patient features
Control group Pred+IFN group
Number of patients Male/female Median age (years) (range) Serum ALT (IU/I)* Serum HBV-DNA (pg/ml)* Median value (range) CPH/CAH
21 14/7 12 (6-15) 102-+56 178-+73 100 (25-800) 17/4
22 15/7 12 (6-16) 100-+53 196-+94 100 (10-800) 17/5
165 D N A and permanent conversion to HBeAb. HBsAg clearance occurred 3 months after the initiation of I F N in one patient and 2 months after I F N withdrawal in the other. Seven (32%) patients in the treated group showed a significant decline in serum H B V - D N A levels of more than 75% of the initial value, which persisted during treatment. Five o f these patients had a peak elevation of A L T (6- to 12-fold) similar to that observed in full responder patients. In one of them, H B V - D N A became negative at the 12th m o n t h o f I F N treatment, but was again detectable in the post-treatment period. These seven patients had persistent H B e A g throughout the observation period and were classified as partial responders. According to the pre-established criteria, the remaining six patients were considered non-responders. O f these, one showed an increase in A L T levels (16x upper normal levels) at the first month o f I F N administration, followed by a decline in A L T values to 2 x the upper normal levels. The other five non-responders did not show any peak activity in A L T throughout the study period. Serum A L T levels were normal at the end o f the study in 13 patients, nine with full response and four with partial response. In the control group, two cases o f HBeAg seroconversion occurred at the 10th and 18th month of observation, respectively. In the first case, H B e A g clearance was preceded by a 6-fold increase in A L T values. N o n e of the untreated patients showed H B s A g clearance during the observation period. The predictive values for a full response o f the baseline levels of H B V - D N A and A L T , of histology and o f age TABLE 3 Predictive value of baseline HBV-DNA and ALT levels, of histology and of age in the response to treatment
* Mean+s.d. CPH=chronic persistent hepatitis; CAH=chronic active hepatitis. Number of patients
Untreated controls
Prednisone plus interferon
21
22
TABLE 2
Responses~total
Response to treatment of the study population
Serum HBV-DNA (pg/ml)* 10-100 > 100
1/11 1/10
9/12 0/10
Serum ALT (IU/1)** <100 > 100
0/12 2/9
7/14 2/8
Histology** CPH CAH
0/17 2/4
8/17 1/5
Age** < 12 years > 12 years
0/I 1 2/10
5/9 4/13
Characteristics
Untreated controls
Prednisone plus interferon
Number of patients Loss of serum HBV-DNA and HBeAg Loss of serum HBsAg
21 2 (9.5%) (1.2-30.4)** 0 (0.0--15.4) 0 (0.(~15_4) 19 (90.5%) (69.6-98.8)
22 9 (41%)* (20.7-63.4) 2 (9%) (1.1-29.2) 7 (32%) (13.9-54_9) 6 (27%) (10.7-50.2)
Serum HBV-DNA decrease*** Serum HBV-DNA unchanged
* p=0.020 vs untreated controls, by the Fisher exact test. ** 95% confidence limits. *** at least 75% of the initial value_
* p<0.05 by chi-square test for heterogeneity_ **p>0.10. CPH= chronic persistent hepatitis; CAH=chronic active hepatitis.
166 were evaluated (Table 3). With the chi-square test for heterogeneity, only baseline HBV-DNA levels were a significant independent predictor of response. In fact, of the 12 patients with a baseline level of HBV-DNA equal to or below the median value of 100 pg/ml, nine (75%) responded to the treatment. The other three showed either a temporary clearance or a significant decrease in serum HBV-DNA during treatment. In contrast, none of the ten patients whose baseline HBV-DNA levels were above 100 pg/ml responded to treatment (p<0.05). In this latter group, four patients showed a decrease in serum HBVDNA levels during treatment. The initial level of HBVDNA in the two untreated patients who had seroconversion to HBeAb was higher than 100 pg/ml in one case and lower in the other. Of the 14 patients with a low (less than 100 U/l) ALT value on entry, seven responded to treatment, while of the eight patients with baseline levels above 100 U/l, a response was observed in two (p>0.10). In the controls, the two responsive patients had an ALT baseline of over 100 U/1. Response to treatment was observed in one of the five patients with chronic active hepatitis and in eight of the 17 patients with chronic persistent hepatitis. Among the controls, the two patients who showed HBeAg seroconversion had chronic active hepatitis. Among the patients with a history of acute type B hepatitis (5 to 9 years earlier) three out of four treated patients responded to therapy. One control patient who had had acute hepatitis 6 years earlier did not show any change during the observation period. Of the seven treated patients who were presumably infected at birth (born to HBsAg positive mothers), only one responded to therapy. The sequential treatment with prednisone and interferon was generally well tolerated. All patients experienced a moderate febrile reaction during the first month of treatment with IFN. Two patients had fever after each dose of IFN, accompanied by headache and arthromyalgia. These patients showed no response to treatment and IFN was discontinued at the 8th month. They were followed up for an additional 12 months and were considered as non-responders in the treated group. All other patients completed the treatment period and the subsequent 6 months of follow up.
Discussion The results of this study demonstrated that the administration of a short course of prednisone followed by a long course of interferon may be an effective and safe regimen in the treatment of children with chronic hepa-
R. UTILI et al. titis B. However, the most relevant finding was that this regimen appeared to be effective only in patients who had low basal levels of viral replication as expressed by HBVDNA serum levels of less than 100 pg/ml. In this group, a high response rate to treatment was observed with 75% permanent clearance of HBV-DNA, seroconversion to anti-HBe and, in two cases, clearance of HBsAg (Table 2). In contrast, the same regimen was ineffective for children with elevated levels of viral replication (baseline HBV-DNA greater than 100 pg/ml). Indeed none of these children cleared HBV-DNA or HBeAg, although 40% of them showed a decline in HBV-DNA serum levels during treatment. The predictive role of the HBV-DNA level before treatment was also demonstrated in adult patients with chronic hepatitis B by Perrillo et al. (4) and by Brook et al. (13). In Perrillo's study, the interferon treatment was more effective in patients with low baseline levels of HBVD N A (less than 100 pg/ml) than in those with an initial level above 200 pg/ml. These results may explain the relatively low efficacy of interferon in a non-selected pediatric population observed in earlier trials, particularly in the Chinese studies (6-8). In fact, most of the Chinese children may have acquired the infection by vertical transmission and may have developed an immunological "tolerance" to HBV, characterized by high viral replication and low disease activity. Under these conditions, both interferon alone and the sequential treatment with prednisone and interferon seem to have a low efficacy in stimulating the immune response against HBV (8). In contrast, children with low levels of viral replication seem to behave immunologically like adult patients and, similarly, may benefit from IFN treatment. These results do not indicate any significant predictive role for basal aminotransferase levels on the outcome of IFN treatment. Earlier studies have generated contrasting data on this point, reporting a positive response to IFN treatment associated with either low or high ALT serum levels (4,5,7). Our data, however, suggest that the marked rebound of ALT, induced by the withdrawal of prednisone, may be an important factor in predicting a positive response to treatment. Most of the patients in this study had chronic persistent hepatitis. It is debatable whether the mild activity of their condition warrants the use of interferon treatment. Yet, these patients may progress to a more active disease (14) and acquire delta superinfection. Considering the potentially favourable response to treatment of patients with low basal viral replication (75% HBeAg clearance and ca. 17% HBsAg clearance) and of the mild side-effects ob-
PREDNISONE-IFN IN CHILDHOOD HEPATITIS B
served, we strongly favour antiviral treatment even in the presence of chronic persistent hepatitis. This trial did not include a group of patients who received only I F N because of the poor, non-significant response to I F N alone observed in published trials (6,7). Therefore, at present it is not possible to establish whether sequential treatment with prednisone and interferon is more effective in children with low basal H B V - D N A levels than I F N alone. Nevertheless, previous data indicate that I F N alone (3 M U thrice weekly) is of limited benefit to children with chronic hepatitis B (6). In a recent study, Ruiz-Moreno et al. (15) showed that a high dose of I F N (10 M U / m 2 three times a week, for 6 months), but not a lower dose (5 MU/m2), was more effective than no treatment in inducing H B V - D N A clearance and HBeAg seroconversion in children with chronic hepatitis B. The effective dose of 10 M U / m 2 is much higher than the dose used in our study. The treatment schedule used in our study may be considered as an alternative to I F N alone at 10 M U / m 2 as proposed by Ruiz-Moreno et al., since it may be better tolerated. A n o t h e r interesting observation in the present study is that seroconversion to anti-HBe and to anti-HBs occurred on several occasions later on in the treatment and even after its discontinuation. This might support prolonged I F N administration. Our conclusion is that a short course of prednisone followed by the administration of 3 M U / m 2 of IFN, three times weekly, for 12 months, is an effective and safe regimen for children with chronic hepatitis B with a basal H B V - D N A serum level lower than 100 pg/ml. For children with higher levels of viral replication, the combined treatment seems to be ineffective and alternative therapeutic regimens are warranted.
Acknowledgements This work was partly supported by CNR, F A T M A project n. 92.00106.PF41 and partly by M U R S T , N a t i o n -
167
al Project on "Liver Cirrhosis and Viral Hepatitis". We thank Prof. Ciro Gallo for his statistical advice.
References 1_ Lok ASF, Lai CL. A longitudinal follow-up of asymptomatic hepatitis B surface antigen-positive Chinese children. Hepatology 1988; 82:1130-2. 2. Bortolotti F, Calzia R, Vegnente A, et al. Chronic hepatitis in childhood: the spectrum of the disease_ Gut 1988; 29: 659-64. 3. Wu TC, Tong MJ, Hwang B, et al. Primary hepatocellular carcinoma and hepatitis B infection during childhood. Hepatology 1987; 7: 46--8. 4_ Perrillo RP, Schiff ER, Gary LD, et al. A randomized, controlled trial of interferon alfa-2b alone and after prednisone withdrawal for the treatment of chronic hepatitis B. N Engl J Med 1990; 323: 295-301. 5. Hoofnagle JH, Peters M, Mullen KD, et al. Randomized, controlled trial of recombinant human alpha-interferon in patients with chronic hepatitis B. Gastroenterology 1988; 95: 1318-25. 6. Utili R, Sagnelli E, Galanti B, et al. Prolonged treatment of children with chronic hepatitis B with recombinant alfa-2a interferon: a controlled, randomized study. Am J Gastroenterol 1991; 86: 327-30. 7. Lai CL, Lok ASF, Lin HI, et al. Placebo-controlled trial of recombinant alfa-2 interferon in Chinese HBsAg-carrier children. Lancet 1987; ii: 877-80. 8. Lok ASF, Lai CL, Wu PC, et al. Alfa-interferon treatment in Chinese patients with chronic hepatitis B_ J Hepatol 1990; I1: SI21-5. 9_ Perrillo RP, Regenstein FG, Peters MG, et al. Prednisone withdrawal followed by recombinant alfa interferon in the treatment of chronic type B hepatitis. Ann Intern Med 1988; 109: 95-100. 10. Perrillo RP. Treatment of chronic hepatitis B with interferon: experience in western countries. Semin Liver Dis 1989;9: 240-8. 11. Nardiello S, Gargiulo M, Pizzella T, et al. Relation of changes in hepatitis B virus replication markers to the outcome of interferon treatment in patients with chronic hepatitis. J Chemother 1992; 4: 95-8. 12. Yusuf S, Peto R, Lewis JA, et al. Beta-blockade during and after myocardial infarction. Prog Cardiovasc Dis 1985; 27: 335-71. 13. Brook MG, Karayiannis P, Thomas HC. Which patients with chronic hepatitis B virus infection will respond to alfa-interferon therapy? A statistical analysis of predictive factors. Hepatology 1989; 10: 761-3_ 14. Giusti G, Galanti B, Gaeta GB, et al. Clinical presentation and natural history of chronic persistent hepatitis. A multicenter retrospective study on 1197 cases. Ital J Gastroenterol 1991; 23:111-8. 15. Ruiz-Moreno M, Rua MJ, Molina J, et al. Prospective, randomized controlled trial of interferon-alfa in children with chronic hepatitis B_ Hepatology 1991; 13: 1035-9.
Copyright © Journalof Hepatology 1994
Journal of Hepatology ISSN 0168-8278
Treatment of chronic hepatitis B in children with prednisone followed by alfa-interferon: a controlled randomized study R i c c a r d o Utili t, Evangelista Sagnelli 2, G i o v a n n i B. G a e t a 2, B r u n o Galanti 2, Salvatore Nardiello 2, Francesca M. Felaco 2, Giuseppe Pasquale 2, Aldo M a r r o n e ~, Lucia A p r e a 2, Teresa Pizzella 2, L a u r a Digilio% Giuseppe Cesaro', Carolina S a r d a r o 2, Lucio Santarpia ~, Luigi E. Adinolfi', Augusto Andreana% Michele Russo 2, Pietro Filippini 2, R o s a Z a m p i n o l, Felice Piccinino 3, Giuseppe Ruggiero t and Guiseppe Giusti 2 Ilnstitute of Medical Therapy, :Institute of lnfectious Diseases and Jlnstitute of Tropical and Subtropical Diseases, University of Naples, 2nd Medical School, Naples, Italy
(Received 10 April 1992)
The efficacy and safety of sequential treatment with prednisone and interferon was evaluated in a randomized, controlled study on 43 children with biopsy proven HBsAg/HBeAg/hepatitis B virus-DNA positive, anti-delta negative, chronic hepatitis (34 chronic persistent hepatitis, 9 chronic active hepatitis). Patients received either a 1-month course of prednisone (0.6 to 0.3 mg/kg per day) followed by interferon alfa-2a (3 MU/m% thrice weekly, for 12 months; 22 patients) or no treatment (21 patients). At the end of the study (20 months), clearance of hepatitis B virus-DNA and HBeAg seroconversion were observed in nine (41%) of the patients treated with prednisone and interferon and in two (9.5%) of the untreated controls (p=0.020). Two of the treated patients who lost HBeAg, also cleared HBsAg. In the treated group, 13 (59%) patients had stable normal levels of alanine aminotransferase on their last examination. The baseline serum level of hepatitis B virus-DNA was an important predictor of response. In fact, HBeAg clearance was observed in 75% of patients with a baseline hepatitis B virusDNA level lower than 100 pg/ml and in none with a level above 100 pg/ml. We suggest that combined treatment with prednisone followed by alfa-interferon may be safe and effective in inducing a stable clearance of HBeAg and, in some cases, of HBsAg in children with chronic hepatitis B and with a low level of viral replication. For children with high levels of viral replication, this regimen seems to be ineffective. © Journal of Hepatology. Key words: Childhood hepatitis B; Chronic hepatitis B; Interferon; Prednisone withdrawal
Chronic hepatitis B in children is usually an asymptomatic disease, associated in most cases with a high level of hepatitis B virus (HBV) replication and low inflammatory activity (1). Nevertheless, these children may experience a flare-up of the diseases which may induce a rapid progression to a more severe form of chronic liver disease and in some cases hepatocellular carcinoma (2,3). Therefore, antiviral treatment for this condition appears warranted. The aim of treatment with alfa interferon (IFN) in these patients is to reduce viral replication and to induce a moderate activation of the immune response to HBV, which may lead to a self-limiting necrosis of the infected
hepatocytes. Consequently, a stable suppression of viral replication (clearance of HBeAg and HBV-DNA) may be expected and, in a few cases, a complete resolution of HBV infection (clearance of HBsAg) may follow. IFN treatment seems to be effective for adults with chronic hepatitis B (4,5). The same treatment, however, seems to be less effective for children (6,7)_ This has been attributed to an immunological tolerance to the virus which, in most cases, was acquired at an early age when the immune system was immature (8). It has been proposed that the response rate to IFN may be enhanced by a priming course of corticosteroids (9).
Correspondence to: Prof. R. Utili, Istituto di Terapia Medica, I Facoltfi di Medicina, UniversitY.di Napoli, Via Cotugno 1 (c/o Ospedale Ges/J e Maria), 80135 Naples, Italy
164 The rationale for this approach is based on the fact that the withdrawal of corticosteroid therapy may induce an immunological rebound, followed by an episode of liver cell necrosis and by a decrease in viral replication (10). This treatment regimen was shown to have a slight advantage over IFN alone in inducing a state of viral suppression in Chinese children (8). There are no similar studies in Caucasian children. Accordingly, we initiated a randomized controlled study to evaluate the efficacy of sequential treatment with a short course of prednisone followed by a long course of IFN in young patients with chronic hepatitis and active HBV replication.
Patients and Methods
Criteria for inclusion in the study were: age below 16 years, infection by HBV with markers of active viral replication in serum (both HBeAg and HBV-DNA) and persistent elevation of serum aminotransferase values (at least 1.5 times the upper normal limit) for at least 6 months; evidence of chronic hepatitis on liver biopsy performed in the preceding 3 months. Exclusion criteria were: history of acute hepatitis in the previous 24 months, previous antiviral or immunosuppressive therapy, presence of antibodies to hepatitis delta virus (HDV), hepatitis C virus (HCV), or human immunodeficiency virus (HIV), decompensated liver disease, white-cell counts below 3x 109 per liter, platelet counts below 70x 109 per liter, drug addiction, or any other serious illnesses. According to these criteria, 43 Italian patients, 6-16 years old, entered the study between September 1989 and November 1990. They were randomly assigned to two groups according to a simple randomization list prepared in advance from a random digit table. Consecutively numbered sealed envelopes with the treatment to be assigned were used to preserve the unpredictability of the next assignment. The first group consisted of 21 patients who were left untreated and served as controls. The second group consisted of 22 patients who received a lmonth course of prednisone, 0.6 mg/kg per day orally for the first 3 weeks and 0.3 mg/kg per day in the 4th week, followed by a 2-week rest and then, treatment with interferon alfa-2a 3 MU/m 2 three times weekly for 12 months. All patients were examined at 2-week intervals during the month of prednisone treatment and during the first 2 months of interferon treatment. Then they were examined at 1-month intervals during the interferon treatment and for the next 6 months after IFN discontinuation; the total follow-up was at least 20 months. Control patients were also evaluated at monthly intervals. On each examination, the patients received a clinical, biochemical, hematolog-
R. UTILI et al. ical and virological assessment and any adverse reactions or side effects of treatment were registered. Interferon dosage was reduced or discontinued in the event of serious side effects. Informed or written consent was obtained from a parent of each patient. HBsAg, HBsAb, HBeAg, HBeAb, and anti-HDV were tested with commercial enzyme-linked immunosorbent assays (Abbott). HCV-Ab was tested by Ortho-HCV ELISA Test System 2nd generation and HIV-Ab by HIV-1 Recombinant (Abbott). The HBV-DNA was detected by slot-hybridization using a 32P-labelled cloned HBV-DNA probe as previously described (6). Semi-quantitation of HBV-DNA was obtained by comparing the autoradiographic signals of appropriately diluted serum samples to those of serial dilutions of a standard serum with known concentrations of HBV-DNA (Genostics Hepatititis B Virus D N A positive control, Abbott). Test sensitivity was 1 pg of HBV-DNA per ml of serum (11). The outcome of the disease was assessed at the end of the observation period. Response to treatment was defined as follows: 1) full, in the presence of irreversible disappearance of HBV-DNA during therapy associated with a stable HBeAg/HBeAb seroconversion which persisted for at least 6 months after the end of therapy, with or without HBsAg clearance; 2) partial, when a loss or at least a stable 75% decrease in serum HBV-DNA concentration occurred during therapy, with persistence of HBeAg; 3) absent, when no change in serum HBV-DNA levels was observed during treatment. No second liver biopsy was performed at the end of the study for ethical reasons. Statistics The results are expressed as mean---SD or, when indicated, as the median value. Analysis of HBV-DNA levels was made after a square-root transformation to achieve normally distributed data. The percentage differences were analyzed by the Fisher exact test. The independent predictors of response were evaluated by the chi-square test for heterogeneity (12).
Results
All 43 patients were asymptomatic on admission and their disease was generally mild, as demonstrated by the predominance of chronic persistent hepatitis vs chronic active hepatitis (Table l). Fifteen (eight controls, seven treated) were born to HBsAg-positive mothers and, presumably, they had acquired the infection by vertical transmission. The other 28 patients belonged to families with at least one relative, other than the mother, infected by HBV. Five of them (one control and four
PREDNISONE-IFN IN CHILDHOOD HEPATITIS B treated) had a past history o f s y m p t o m a t i c acute viral hepatitis. The control and treated groups were c o m p a r a b l e with respect to age, sex ratio, biochemical and histologic features and baseline H B V - D N A (Table 1). According to the study protocol, a full response (permanent clearance o f H B V - D N A and H B e A g seroconversion) was observed in nine (41%) patients treated with prednisone plus interferon, and in two (9.5%) of the untreated controls (p=0.020 by the Fisher exact test). A m o n g treated patients who responded to treatment, two also cleared HBsAg. All patients who responded to treatment had a peak elevation o f alanine aminotransferase (ALT) activity (from 4 to 25 times upper n o r m a l levels) within the 2rid m o n t h of treatment with interferon, and cleared H B V - D N A shortly thereafter_ In eight responders, seroconversion to H B e A b occurred from 1 to 9 months after the initiation o f I F N treatment. The ninth responder patient seroconverted to H B e A b 2 months after treatment was discontinued. In one patient who had shown H B e A g seroconversion early on in I F N treatment, H B e A g and H B V - D N A reappeared at the end o f I F N administration. However, 2 months later he had an increase in A L T values with a stable clearance of HBV-
TABLE I Characteristics of patients on admission Patient features
Control group Pred+IFN group
Number of patients Male/female Median age (years) (range) Serum ALT (IU/I)* Serum HBV-DNA (pg/ml)* Median value (range) CPH/CAH
21 14/7 12 (6-15) 102-+56 178-+73 100 (25-800) 17/4
22 15/7 12 (6-16) 100-+53 196-+94 100 (10-800) 17/5
165 D N A and permanent conversion to HBeAb. HBsAg clearance occurred 3 months after the initiation of I F N in one patient and 2 months after I F N withdrawal in the other. Seven (32%) patients in the treated group showed a significant decline in serum H B V - D N A levels of more than 75% of the initial value, which persisted during treatment. Five o f these patients had a peak elevation of A L T (6- to 12-fold) similar to that observed in full responder patients. In one of them, H B V - D N A became negative at the 12th m o n t h o f I F N treatment, but was again detectable in the post-treatment period. These seven patients had persistent H B e A g throughout the observation period and were classified as partial responders. According to the pre-established criteria, the remaining six patients were considered non-responders. O f these, one showed an increase in A L T levels (16x upper normal levels) at the first month o f I F N administration, followed by a decline in A L T values to 2 x the upper normal levels. The other five non-responders did not show any peak activity in A L T throughout the study period. Serum A L T levels were normal at the end o f the study in 13 patients, nine with full response and four with partial response. In the control group, two cases o f HBeAg seroconversion occurred at the 10th and 18th month of observation, respectively. In the first case, H B e A g clearance was preceded by a 6-fold increase in A L T values. N o n e of the untreated patients showed H B s A g clearance during the observation period. The predictive values for a full response o f the baseline levels of H B V - D N A and A L T , of histology and o f age TABLE 3 Predictive value of baseline HBV-DNA and ALT levels, of histology and of age in the response to treatment
* Mean+s.d. CPH=chronic persistent hepatitis; CAH=chronic active hepatitis. Number of patients
Untreated controls
Prednisone plus interferon
21
22
TABLE 2
Responses~total
Response to treatment of the study population
Serum HBV-DNA (pg/ml)* 10-100 > 100
1/11 1/10
9/12 0/10
Serum ALT (IU/1)** <100 > 100
0/12 2/9
7/14 2/8
Histology** CPH CAH
0/17 2/4
8/17 1/5
Age** < 12 years > 12 years
0/I 1 2/10
5/9 4/13
Characteristics
Untreated controls
Prednisone plus interferon
Number of patients Loss of serum HBV-DNA and HBeAg Loss of serum HBsAg
21 2 (9.5%) (1.2-30.4)** 0 (0.0--15.4) 0 (0.(~15_4) 19 (90.5%) (69.6-98.8)
22 9 (41%)* (20.7-63.4) 2 (9%) (1.1-29.2) 7 (32%) (13.9-54_9) 6 (27%) (10.7-50.2)
Serum HBV-DNA decrease*** Serum HBV-DNA unchanged
* p=0.020 vs untreated controls, by the Fisher exact test. ** 95% confidence limits. *** at least 75% of the initial value_
* p<0.05 by chi-square test for heterogeneity_ **p>0.10. CPH= chronic persistent hepatitis; CAH=chronic active hepatitis.
166 were evaluated (Table 3). With the chi-square test for heterogeneity, only baseline HBV-DNA levels were a significant independent predictor of response. In fact, of the 12 patients with a baseline level of HBV-DNA equal to or below the median value of 100 pg/ml, nine (75%) responded to the treatment. The other three showed either a temporary clearance or a significant decrease in serum HBV-DNA during treatment. In contrast, none of the ten patients whose baseline HBV-DNA levels were above 100 pg/ml responded to treatment (p<0.05). In this latter group, four patients showed a decrease in serum HBVDNA levels during treatment. The initial level of HBVDNA in the two untreated patients who had seroconversion to HBeAb was higher than 100 pg/ml in one case and lower in the other. Of the 14 patients with a low (less than 100 U/l) ALT value on entry, seven responded to treatment, while of the eight patients with baseline levels above 100 U/l, a response was observed in two (p>0.10). In the controls, the two responsive patients had an ALT baseline of over 100 U/1. Response to treatment was observed in one of the five patients with chronic active hepatitis and in eight of the 17 patients with chronic persistent hepatitis. Among the controls, the two patients who showed HBeAg seroconversion had chronic active hepatitis. Among the patients with a history of acute type B hepatitis (5 to 9 years earlier) three out of four treated patients responded to therapy. One control patient who had had acute hepatitis 6 years earlier did not show any change during the observation period. Of the seven treated patients who were presumably infected at birth (born to HBsAg positive mothers), only one responded to therapy. The sequential treatment with prednisone and interferon was generally well tolerated. All patients experienced a moderate febrile reaction during the first month of treatment with IFN. Two patients had fever after each dose of IFN, accompanied by headache and arthromyalgia. These patients showed no response to treatment and IFN was discontinued at the 8th month. They were followed up for an additional 12 months and were considered as non-responders in the treated group. All other patients completed the treatment period and the subsequent 6 months of follow up.
Discussion The results of this study demonstrated that the administration of a short course of prednisone followed by a long course of interferon may be an effective and safe regimen in the treatment of children with chronic hepa-
R. UTILI et al. titis B. However, the most relevant finding was that this regimen appeared to be effective only in patients who had low basal levels of viral replication as expressed by HBVDNA serum levels of less than 100 pg/ml. In this group, a high response rate to treatment was observed with 75% permanent clearance of HBV-DNA, seroconversion to anti-HBe and, in two cases, clearance of HBsAg (Table 2). In contrast, the same regimen was ineffective for children with elevated levels of viral replication (baseline HBV-DNA greater than 100 pg/ml). Indeed none of these children cleared HBV-DNA or HBeAg, although 40% of them showed a decline in HBV-DNA serum levels during treatment. The predictive role of the HBV-DNA level before treatment was also demonstrated in adult patients with chronic hepatitis B by Perrillo et al. (4) and by Brook et al. (13). In Perrillo's study, the interferon treatment was more effective in patients with low baseline levels of HBVD N A (less than 100 pg/ml) than in those with an initial level above 200 pg/ml. These results may explain the relatively low efficacy of interferon in a non-selected pediatric population observed in earlier trials, particularly in the Chinese studies (6-8). In fact, most of the Chinese children may have acquired the infection by vertical transmission and may have developed an immunological "tolerance" to HBV, characterized by high viral replication and low disease activity. Under these conditions, both interferon alone and the sequential treatment with prednisone and interferon seem to have a low efficacy in stimulating the immune response against HBV (8). In contrast, children with low levels of viral replication seem to behave immunologically like adult patients and, similarly, may benefit from IFN treatment. These results do not indicate any significant predictive role for basal aminotransferase levels on the outcome of IFN treatment. Earlier studies have generated contrasting data on this point, reporting a positive response to IFN treatment associated with either low or high ALT serum levels (4,5,7). Our data, however, suggest that the marked rebound of ALT, induced by the withdrawal of prednisone, may be an important factor in predicting a positive response to treatment. Most of the patients in this study had chronic persistent hepatitis. It is debatable whether the mild activity of their condition warrants the use of interferon treatment. Yet, these patients may progress to a more active disease (14) and acquire delta superinfection. Considering the potentially favourable response to treatment of patients with low basal viral replication (75% HBeAg clearance and ca. 17% HBsAg clearance) and of the mild side-effects ob-
PREDNISONE-IFN IN CHILDHOOD HEPATITIS B
served, we strongly favour antiviral treatment even in the presence of chronic persistent hepatitis. This trial did not include a group of patients who received only I F N because of the poor, non-significant response to I F N alone observed in published trials (6,7). Therefore, at present it is not possible to establish whether sequential treatment with prednisone and interferon is more effective in children with low basal H B V - D N A levels than I F N alone. Nevertheless, previous data indicate that I F N alone (3 M U thrice weekly) is of limited benefit to children with chronic hepatitis B (6). In a recent study, Ruiz-Moreno et al. (15) showed that a high dose of I F N (10 M U / m 2 three times a week, for 6 months), but not a lower dose (5 MU/m2), was more effective than no treatment in inducing H B V - D N A clearance and HBeAg seroconversion in children with chronic hepatitis B. The effective dose of 10 M U / m 2 is much higher than the dose used in our study. The treatment schedule used in our study may be considered as an alternative to I F N alone at 10 M U / m 2 as proposed by Ruiz-Moreno et al., since it may be better tolerated. A n o t h e r interesting observation in the present study is that seroconversion to anti-HBe and to anti-HBs occurred on several occasions later on in the treatment and even after its discontinuation. This might support prolonged I F N administration. Our conclusion is that a short course of prednisone followed by the administration of 3 M U / m 2 of IFN, three times weekly, for 12 months, is an effective and safe regimen for children with chronic hepatitis B with a basal H B V - D N A serum level lower than 100 pg/ml. For children with higher levels of viral replication, the combined treatment seems to be ineffective and alternative therapeutic regimens are warranted.
Acknowledgements This work was partly supported by CNR, F A T M A project n. 92.00106.PF41 and partly by M U R S T , N a t i o n -
167
al Project on "Liver Cirrhosis and Viral Hepatitis". We thank Prof. Ciro Gallo for his statistical advice.
References 1_ Lok ASF, Lai CL. A longitudinal follow-up of asymptomatic hepatitis B surface antigen-positive Chinese children. Hepatology 1988; 82:1130-2. 2. Bortolotti F, Calzia R, Vegnente A, et al. Chronic hepatitis in childhood: the spectrum of the disease_ Gut 1988; 29: 659-64. 3. Wu TC, Tong MJ, Hwang B, et al. Primary hepatocellular carcinoma and hepatitis B infection during childhood. Hepatology 1987; 7: 46--8. 4_ Perrillo RP, Schiff ER, Gary LD, et al. A randomized, controlled trial of interferon alfa-2b alone and after prednisone withdrawal for the treatment of chronic hepatitis B. N Engl J Med 1990; 323: 295-301. 5. Hoofnagle JH, Peters M, Mullen KD, et al. Randomized, controlled trial of recombinant human alpha-interferon in patients with chronic hepatitis B. Gastroenterology 1988; 95: 1318-25. 6. Utili R, Sagnelli E, Galanti B, et al. Prolonged treatment of children with chronic hepatitis B with recombinant alfa-2a interferon: a controlled, randomized study. Am J Gastroenterol 1991; 86: 327-30. 7. Lai CL, Lok ASF, Lin HI, et al. Placebo-controlled trial of recombinant alfa-2 interferon in Chinese HBsAg-carrier children. Lancet 1987; ii: 877-80. 8. Lok ASF, Lai CL, Wu PC, et al. Alfa-interferon treatment in Chinese patients with chronic hepatitis B_ J Hepatol 1990; I1: SI21-5. 9_ Perrillo RP, Regenstein FG, Peters MG, et al. Prednisone withdrawal followed by recombinant alfa interferon in the treatment of chronic type B hepatitis. Ann Intern Med 1988; 109: 95-100. 10. Perrillo RP. Treatment of chronic hepatitis B with interferon: experience in western countries. Semin Liver Dis 1989;9: 240-8. 11. Nardiello S, Gargiulo M, Pizzella T, et al. Relation of changes in hepatitis B virus replication markers to the outcome of interferon treatment in patients with chronic hepatitis. J Chemother 1992; 4: 95-8. 12. Yusuf S, Peto R, Lewis JA, et al. Beta-blockade during and after myocardial infarction. Prog Cardiovasc Dis 1985; 27: 335-71. 13. Brook MG, Karayiannis P, Thomas HC. Which patients with chronic hepatitis B virus infection will respond to alfa-interferon therapy? A statistical analysis of predictive factors. Hepatology 1989; 10: 761-3_ 14. Giusti G, Galanti B, Gaeta GB, et al. Clinical presentation and natural history of chronic persistent hepatitis. A multicenter retrospective study on 1197 cases. Ital J Gastroenterol 1991; 23:111-8. 15. Ruiz-Moreno M, Rua MJ, Molina J, et al. Prospective, randomized controlled trial of interferon-alfa in children with chronic hepatitis B_ Hepatology 1991; 13: 1035-9.