Objective      To investigate the clinical characteristics and outcome of hepatitis B virus (HBV)-related patients’ relapse after nucleoside/nucleotide analogues (NAs) withdrawal and its impact factors.       Methods      The clinical data of HBV-related inpatients of Southwest Hospital in 2011 were retrospect

ively analyzed. The patients who meet the following criteria: 18 to 65 years old

took NAs more than 6 months before withdrawal and admitted patients due to disease onset after NAs withdrawal were recruited

with 152 cases in total. The recovery and exacerbation rate

duration of treatment of prior NAs

time of drug withdrawal to illness onset

time of illness onset to admission

and parameters of serological

biochemical and virological tests after admission were studied.       Results      In the patients

98% withdrew on their own purposes and 82.2% (125/152) recovered and were discharged from hospital

while 17.8% (27/152) were exacerbated

including two dead cases. 17.8% of the patients had liver disease-related complications. 69.1% took NAs for 1 to 3 years before withdrawal. The median time of drug withdrawal to illness onset was 6 months. 74.3% of the patients relapsed within one year after NAs withdrawal. The ALT level in all patients showed a decrease trend after admission

whereas TBIL and PT levels showed parallel changes in each group: two patterns in the recovery group

a steady-then-descending pattern and a progressively-descending pattern

and a progressively-elevating pattern in the exacerbated group. Diagnosed severe hepatitis on admission

shorter time of drug withdrawal to illness onset and higher TBIL level

and prolonged PT were negatively correlated with clinical outcome (P0.05). NAs category

HBeAg state and rebounded HBV DNA level were not significantly relevant to clinical outcome (

P

0.05).       Conclusion      About 1/5 of HBV-related inpatients relapsed after NAs withdrawal have poor clinical outcome. The dynamic changes of TBIL and PT level have some particular patterns

and may be used to predict the clinical outcome. Shorter time of NAs withdrawal to illness onset indicates worse clinical outcome.