Clinical and pathological features in 138 cases of drug-induced liver injury / 中华肝脏病杂志

<p><b>OBJECTIVE</b>To explore the categories of drugs causing hepatotoxicity and analyze the clinical and histological features of the corresponding drug-induced liver injury (DILI), in order to gain insights into potential diagnostic factors for DILI.</p><p><b>METHODS</b>A total of 138 DILI patients treated at our hospital from April 2008 to April 2010 were retrospectively analyzed. The responsible drug for each DILI case was recorded. The Roussel Uclaf Causality Assessment Method (RUCAM) had been used to diagnose DILI. Only cases that had scored as highly probable or probable (more than or equal to 6 points by RUCAM) were included in this study. The patients' general condition, clinical manifestations, and serum biochemical and immunological parameters were assessed. Sixty-six of the patients underwent liver biopsy, and were assessed for liver pathological changes. Clinical and laboratory test data were collected and used to classify the total 138 cases as hepatocellular injury, cholestatic, or mixed hepatocellular-cholestatic types.</p><p><b>RESULTS</b>Within our patient population, the leading cause of DILI was Chinese herb medicine, accounting for 53.62% of cases. Antibiotics were implicated in 7.97% of cases, and dietary supplement in 6.52% of cases. Correlation between the clinical features and histological injury pattern was stronger at the time of biopsy (more than or equal to 3 days after laboratory results) (kappa = 0.63, P less than 0.05) than at the onset of DILI (kappa = 0.25, P less than 0.05). All modified hepatic activity index (HAI) necroinflammatory scores and fibrosis scores were more severe in the cholestatic and mixed injury types than in the hepatocellular injury type (P less than 0.01 and P less than 0.05, respectively).</p><p><b>CONCLUSION</b>Chinese herbal medicine, dietary supplements and antibiotics were the main causes of DILI in our patient population. The clinical and histological features correlated well, especially at later stages of DILI. The degree of inflammation and fibrosis was significantly higher in cholestatic and mixed hepatocellular-cholestatic injury types than in the hepatocellular injury type. Assessment of both clinical and pathological features may represent a more accurate diagnostic method for DILI.</p>

Viral factors influencing histological changes of HBeAg-negative chronic hepatitis B patients with persistently normal serum ALT levels / 中华肝脏病杂志

<p><b>OBJECTIVE</b>To investigate the correlation between viral factors and liver histological changes of HBeAg-negative chronic hepatitis B patients with persistently normal serum ALT levels (PNAL).</p><p><b>METHODS</b>HBV DNA level, HBV genotype, basal core promoter (BCP) and precore mutation were examined in 52 HBeAg-negative chronic hepatitis B patients with PNAL (defined as normal ALT measured on at least 3 occasions in the intervals of about two months over a period of 12 months or more prior to the biopsy).</p><p><b>RESULTS</b>Subjects with both BCP and precore mutations had significantly higher HBV DNA levels than those without mutations [(4.9+/-1.4) vs (4.1+/-1.1) log(10)copies/ml, t = 2.308, P < 0.05]. A higher proportion of patients with histological activity index (HAI) > or = to 4 was found in patients with both mutations (32.1% vs 16.7%) than in patients without mutation, however, the proportion of patients with histological activity index (HAI) > or = to 3 in patients with mutations was not significantly different from that in patients without mutations (14.3% vs. 12.5%, x(2)=0.000, P > 0.05). In patients without precore or BCP mutations, there was a strong positive correlation between viral load and liver inflammation as well as fibrosis (precore: r=0.626, 0.592, P < 0.01; BCP: r=0.730, 0.641, P < 0.01). In patients without both mutations, HBV DNA has shown a high accuracy for predecting fibrosis (F > or = 3) (AUC = 0.905, 95% CI: 0.771-1.039, P < 0.05) with the cutoff value of 4.5 log(10) copies/ml (sensitivity = 1.000, specificity = 0.778, PPV = 42.9%, NPV = 100.0%). Results of both genotypes and mutations were successfully obtained in 40 samples with HBV DNA is > or = to 10(4) copies/ml. The higher viral load was observed in the patients with genotype B than genotype C (5.1 vs 4.3 log(10)copies/ml, t = 2.059, P < 0.05), but no difference was seen of liver pathologic changes between these two genotypes.</p><p><b>CONCLUSIONS</b>Virus harboring both BCP and precore mutants has the higher replication level than wild type virus. 32.1% and 14.3% of the patients with both mutations have moderate or severe inflammation and fibrosis. There was a strong positive correlation between viral load and liver histological changes in patients without precore or BCP mutations, and viral load shows a high accuracy for predecting significant fibrosis (F > or = 3).</p>

FibroTest-ActiTest for predicting liver fibrosis and inflammatory activity in Chinese patients with chronic hepatitis B / 中华肝脏病杂志

<p><b>OBJECTIVE</b>To confirm the diagnostic value of FibroTest-ActiTest (FT-AT) in predicting liver fibrosis and inflammatory activity in patients with chronic hepatitis B (CHB), and to study the discordances between FT and their liver biopsies.</p><p><b>METHODS</b>A study was performed on 100 patients with CHB who underwent liver biopsies in our hospital. Serum samples for biochemical markers were taken on the day of their biopsies. Diagnostic accuracies were assessed by ROC curve analysis.</p><p><b>RESULTS</b>The median biopsy specimen size was 15 mm (range: 8-30), with 9 (median) portal tracts (range 5-26). Thirty-nine patients were classified as Ishak F3-F6 in fibrosis and 65 patients as A2-A4 in inflammation. Areas under ROC curve for diagnosis of significant inflammation (A2-A4), significant fibrosis (F3-F6), and cirrhosis (F5-F6) were 0.833 (95% CI: 0.753-0.913), 0.840 (0.750-0.929), and 0.862 (0.721-1.003), respectively. FT less than 0.31 had a NPV of 86% for excluding significant fibrosis, whereas FT > or = 0.72 had a PPV of 92% for predicting significant fibrosis. Among the 26 patients with 2 fibrosis stages of discordances between FT and biopsy, the discordance was attributable to biopsy in 3 cases, to FT in 7, and undetermined in 16.</p><p><b>CONCLUSION</b>This study confirms the diagnostic value of FT-AT and suggests that 68% of our patients with CHB can be reliably identified by FT without a liver biopsy and with a diagnosis accuracy of 87%.</p>

Histological changes in the livers of chronic hepatitis B patients with persistently normal serum alanine transaminase levels / 中华肝脏病杂志

<p><b>OBJECTIVE</b>To study the histological changes in livers of chronic hepatitis B (CHB) patients with persistently normal serum ALT levels (PNAL).</p><p><b>METHODS</b>274 CHB patients who had percutaneous liver biopsies and had a detectable viral load (lower limit of detection is 10(3) copies/ml) in our department between October 2003 and February 2007 were included in this study. Among these patients, 139 had PNAL, group A, (with at least 3 normal serum ALT levels, with intervals of more than two months over a period of 12 or more months before the biopsy). The other 135 patients, group B, had abnormal serum ALT levels during the same period. The histological changes in the livers of the two groups of patients were compared.</p><p><b>RESULTS</b>Sixty-six (47.5%) patients with PNAL had normal liver histology, but significant pathohistological changes such as significant necroinflammation, fibrosis and/or cirrhosis were found in 33 (23.7%) patients. Thirteen (9.4%) had established cirrhosis. When compared to patients within (0-0.75)x upper limit of normal (ULN) ALT, patients within (0.76-1.00)x ULN ALT had higher scores of histological changes (43.5% vs. 19.8%, P < 0.05). In the PNAL group, scores of histological changes increased sharply in parallel with an age increase of older than 40 yrs. However neither viral loads nor HBeAg statuses of the PNAL patients had any predictive meaning to the scores of the histological findings.</p><p><b>CONCLUSIONS</b>23.7% of our CHB patients with PNAL, regardless of what their HBeAg statuses or viral load levels were, had significant liver pathohistological changes. Liver biopsies should be considered in CHB patients with PNAL, especially those older than 40 yrs and with a higher ALT within (0.76-1) x ULN.</p>