Prednisolone Therapy Accelerated Recovery of Severe Drug-Induced Liver Injury: A Prospective Randomized Controlled Study

Background Drug-induced liver injury (DILI) is one of the most serious adverse drug reactions and the incidence has been increasing rapidly. Accumulating evidence suggested that the immune activation and systemic inammatory responses play a signicant role in the progression of DILI. Corticosteroids are often used in DILI, but clinical usefulness remain controversial. We therefore conducted a prospective randomized controlled study to investigate whether corticosteroid therapy can accelerate recovery and reduce mortality in severe DILI (SDILI). Methods SDILI patients with total bilirubin (TBIL) ≥ 171 μmol/L presented to Fifth Medical Center of PLA General Hospital, Beijing from 1/1/2015 to 31/12/2019 were randomized into prednisolone group and control group. The primary endpoints were proportion of subjects with resolution of SDILI dened as decrease in TBIL by at least 35 μmol/L to below 171 μmol/L and overall survival at 6 months. Patients in prednisolone group received prednisolone 60 mg/day therapy for the rst 7 days. Patients reaching the primary endpoint or achieved decrease in TBIL by more than 35 umol/L on day 8 would continue on tapering doses of prednisolone, otherwise prednisolone would be discontinued. Results On day 8, 50.7% (34/67) and 26.5% (18/68) of the subjects in the prednisolone group and control group achieved the primary endpoint respectively, p=0.002. However, there was no signicant difference in overall survival at 6 months, 95.52% (64/67) vs 91.2% (62/68), p= 0.2. All deaths in both groups occurred in patients who failed to achieve SDILI resolution on day 8. Conclusions Prednisolone therapy may accelerate recovery of SDILI and shorten hospitalization.

Implicated drugs include prescription medications, traditional Chinese medicine (TCM), natural medicine, health product, dietary supplement, and their associated metabolites [1][2][3][4]. Two retrospective studies showed that DILI and drug-induced acute liver failure (ALF) had been increasing rapidly in China and were associated with high mortality [5,6].
The clinical manifestations of DILI can mimic all forms of liver injury, from asymptomatic liver-function abnormalities to acute, subacute, chronic liver disease and even fulminating liver failure [7,8]. Severe DILI (SDILI) with markedly elevated bilirubin can progress rapidly with fatal outcome. Recently, there are major advances in understanding the mechanism of liver injuries, pharmacological properties of the offending drugs and genetic risk models. However, there is little progress in the treatment of SDILI in the last 20 years [8-10]. Discontinuation of the culprit medications and supportive care remain the mainstay of management.
Accumulating evidence suggests that immune activation and systemic in ammatory responses play a signi cant role in the occurrence and progression of DILI [11]. Corticosteroid with its anti-in ammatory effect is often proposed as a treatment option when other managements fail, especially in SDILI with hyperbilirubinemia. Ma et al showed that corticosteroid treatment was associated with a more rapid decline of bilirubin and liver enzymes, but the overall prognosis was not improved because of worsening infectious complications [12]. Other reported trials of corticosteroid therapy demonstrated limited bene ts in DILI [13]. There was reported improvement in drug-induced cholestasis, but data from large scale randomized trial were lacking [14]. Clinical practice guidelines of European Association for the Study of the Liver (EASL) and Asia Paci c Association of Study of Liver Consensus (APASL) do not give clear recommendations about corticosteroids in SDILI.
Therefore, we conducted this prospective randomized controlled study to investigate whether corticosteroid therapy can accelerate recovery and reduce mortality in SDILI treatment.

Study design
This is a randomized control study. The study protocol and procedures were in line with the ethical principles of the

Treatment endpoints
The primary endpoint was: 1) resolution of SDILI de ned as decrease in bilirubin by at least 35 µmol/L to <171µmol/L, and 2) overall survival at 6 months.
Proportion of subjects in the prednisolone and control groups achieving the primary endpoint of resolution of SDILI at the end of week 1, 2, 4, 12 would be compared.

Treatment and follow-up protocol
All patients were given comprehensive medical treatment of liver protection drugs (such as glycyrrhizin and ademetionine).
Patients in prednisolone group received prednisolone 60 mg/day therapy for the rst 7 days which was called implosive therapy period. Patients reaching the primary endpoint or achieving decrease in bilirubin by more than 35 umol/L (2x ULN) on day 8 would continue on tapering doses of prednisolone, otherwise prednisolone would be discontinued.
Prednisolone would be tapered as follows: prednisolone 40mg/day for 7days (consolidation therapy period), 30mg/day for 7 days (remission therapy period), 20mg/day for 7 days (maintenance therapy period), 10 mg/day for 7 days, then off therapy. The ow chart of the study was shown as Figure 1. All patients could receive antibiotics and fresh frozen plasma treatment when indicated.
The dynamic changes of laboratory data were collected at baseline, Week 1, 2, 4 and 12. The survival rates were estimated at Week 4, 12 and 24. All adverse events would be recorded.
Sample size determination Assuming a difference of at least 20% in the primary endpoints to be of clinical interest, we calculated the sample size of each group was 65 cases (α = 0.05 β = 0.2, two-sided test) [20,21].

Statistical Analyses
Data were analyzed with the Statistical Package for Social Sciences (SPSS 12.0). Variables were examined using descriptive statistics. Analysis of variance (ANOVA) was used for comparisons of groups. Nonparametric analyses (Kruskal-Wallis test) were performed when variables did not follow a normal distribution. Categorical data were expressed as proportions and analyzed by χ2 test or the Fisher exact test. Survival curves were plotted by the Kaplan-Meier method and analyzed by the log-rank test. A p-value < 0.05 was considered statistically signi cant.

Results
Baseline characteristics of the study population   The primary endpoint was resolution of severe DILI de ned as decrease in bilirubin by at least 35µmol/L to <171µmol/L.

Survival at 6 months
There was no signi cant difference in overall survival at 6 months between the 2 groups, 95.52% (64/67) vs 91.2% (62/68), p= 0.2, Figure 2. All deaths occurred in patients who failed to achieve resolution of severe DILI on day 8, the death rates among such patients were 9% (3/33) and 12% (6/50) respectively in the prednisolone and control groups, Table 3. Adverse events There was no signi cance in infectious complications, gastrointestinal bleedings and hyperglycemia between the two groups, Table 4. Discussion DILI has become a major health and economic burden. Timely withdrawal of the offending medication is the most important treatment strategy for DILI. Currently, there is no de nitive therapy or approved antidote available for idiosyncratic DILI. Empiric uses of corticosteroid are sometimes given when all other treatments fail, although it is not generally recommended due to the lack of obvious clinical bene t in previously reported studies [22][23][24]. A retrospective analysis of 361 patients with autoimmune ALF, indeterminate ALF and drug-induced ALF suggested that corticosteroids treatment did not improve the overall survival. In the subgroup of patients with drug-induced ALF, the rate of survival was 69% for those who received corticosteroids versus 66% for those who did not (p=0.82) [23]. On the other hand, Czaja et al reported that corticosteroid therapy was effective in female patients with drug induced autoimmune hepatitis [24]. Other studies showed that corticosteroid treatment was associated with a more rapid decline of bilirubin and liver enzymes, but the overall prognosis was not improved [12][13][14]25].
In our studies, about 50% of subjects given prednisolone therapy had resolution of SDILI after 1 week of treatment. All these patients continued to improve and none of them required arti cial hepatic assisted device, liver transplantation or perished at 6 months followed-up. On the other hand, only about 25% of the subjects in the controlled arm had resolution of SDILI by day 8. There was no difference in survival at 6 months, but our sample size did not have the power to detect a small difference in survival. There may be some debates about our de nitions of SDILI and its resolution. We chose TBIL ≥171 µmol/L as our cut-off point because 171 µmol/L is 10x ULN in our laboratory. In order to avoid including those cases that had only minimal decrease of TBIL from just above 171 µmol/L at baseline to just below 171 µmol/L by week 1 as achieving resolution of SDILI, we felt that the decrease must be at least 35 µmol/L (>2x ULN), equivalent to at least 20% of baseline TBIL values. The mean decrease in TBIL in the prednisolone group was 105.1 ± 21.37 µmol/L. For the 34 prednisolone treated patients achieving the primary endpoint, the mean decrease in bilirubin was 128.2 µmol/L, range 48.3 to 245 µmol/L, 22 had decrease of more than 100 µmol/L. The clinical bene ts of prednisolone were observed mainly at the end of the rst week. We discontinued prednisolone therapy in subjects who failed to show a signi cant decrease in TIBIL by day 8 and for responsive subjects, and we tapered the prednisolone dosage over 6 weeks to minimize the side effects of prolonged high dose steroid therapy. Our data suggested that prednisolone therapy, may accelerate recovery of SDILI, allowing early discharge of patients who achieve SDILI resolution and cost savings. The greatest strength of the study is all subjects were managed at one center by the same team with strict adherence to study protocols, variation in supportive care, missing data, loss to follow-ups and protocol violations were not issues at all. On the other hand, one may argue that whether the experience at one center was applicable to other centers. A larger multi-center trial may address the above limitations.

Declarations
Availability of data and material The datasets generated and analysed during the current study cannot be available for public access due to the privacy of patients, but can be obtained from the corresponding author on reasonable request approved by the Ethics Committee of Fifth Medical Center of Chinese PLA General Hospital.

Ethics approval and clinical trial registration
The study protocol was approved by the Ethics Committee of Fifth Medical Center of Chinese PLA General Hospital. The trial was registered at the Chinese Clinical Trial Registry with identi er ChiCTR-IOR-17010880.
Informed consent to participates and publish The ow chart of the study Figure 2 Overall survival curve analysis at 24W between prednisolone and control groups. There were no signi cant differences in 24W survival between the two groups (p=0.2).