Early View (Online Version of Record published before inclusion in an issue)
Article has an altmetric score of 1
Summary
Background
Both spontaneous and nucleos(t)ide analogue (Nuc)-treated hepatitis B surface antigen (HBsAg) seroclearance are associated with excellent clinical outcomes.
Aim
To conduct a case–control study to explore whether there is difference of clinical outcomes between these two groups.
Methods
A total of 312 chronic hepatitis B patients with spontaneous HBsAg seroclearance and 110 patients with Nuc-treated HBsAg seroclearance were recruited retrospectively. Propensity score (PS) matching method produced 98 patients in each group for comparison. The development of hepatocellular carcinoma (HCC), hepatic complications and cumulative incidence of antibody to HBsAg (anti-HBs) was compared.
Results
During a mean follow-up period of 107 months after HBsAg seroclearance, five patients developed HCC after a mean period of 75.3 months (four and one patients with spontaneous and Nuc-treated HBsAg seroclearance, respectively) in overall population. One died of pneumonia with sepsis and one experienced variceal bleeding in Nuc-treated patients but none in spontaneous group. The incidence of anti-HBs seroconversion was comparable between spontaneous and Nuc-treated HBsAg seroclearance (69.6% vs. 66.4%, respectively, P = 0.617). There were no significant differences in HCC development (2% vs. 1.1%), overall mortality (0% vs. 1%), variceal bleeding (0% vs. 4.2%) and 6-year cumulative incidence of anti-HBs seroconversion (62.3% vs. 61.5%) among PS-matched patients with spontaneous and Nuc-treated HBsAg seroclearance.
Conclusions
The clinical outcomes between patients with spontaneous and Nuc-treated HBsAg seroclearance are comparable. HCC can develop at a low rate during long-term follow-up and periodic surveillance after HBsAg seroclearance is still mandatory. 作者: StephenW 时间: 2016-4-15 15:51
后自发和核苷（酸）类似物IDE治疗乙肝表面抗原血清学清除慢性乙型肝炎临床转归
Y. C. Chen1,2，* W. J. Jeng1,2，R. N. Chien3，C. M. Chu1,2安迪。 F. Liaw1
HCC development and hepatic complications after HBsAg seroclearance
Of the 312 patients with spontaneous HBsAg seroclearance, four male patients developed HCC during follow-up. A 49-year-old noncirrhotic, genotype C patient developed HCC 61.5 months after HBsAg seroclearance. The remaining three patients were 52, 53 and 63 years of age and had evidence of cirrhosis at the time of HBsAg seroclearance (2 genotype B and 1 undetermined) and HCC occurred 110.2, 117.2 and 31 months, respectively, after HBsAg seroclearance. Of the 110 Nuc-treated patients with HBsAg seroclearance, a 59-year-old noncirrhotic, genotype B male patient developed HCC 56.7 months after HBsAg seroclearance. The estimated annual incidence of HCC development was 0.13% and 0.14% in spontaneous and Nuc-treated groups, respectively (P = 0.906). The estimated annual incidence of HCC development was corrected with SIR and the corresponding figures were 0.13% and 0.16%, respectively. In Nuc-treated group, one patient died of pneumonia with sepsis and another patient with pre-existing cirrhosis experienced oesophageal variceal bleeding during follow-up. None of patients in spontaneous group died or had subsequent hepatic complication. Among the PS-matched patients, there was no significant difference in HCC development (2% vs. 1.1%), overall mortality (0% vs. 1%) and variceal bleeding episode (0% vs. 4.2%) (P = 1.000) (Table 2). The estimated annual incidence of HCC was comparable between spontaneous and Nuc-treated groups (0.31% vs. 0.16%, P = 0.735).
After excluding the 44 patients with cirrhosis at the time of HBsAg seroclearance, one of 290 patients with spontaneous HBsAg seroclearance and one of 88 Nuc-treated patients with HBsAg seroclearance developed HCC during a mean follow-up period of 117.5 and 79 months, respectively. The estimated annual incidence of HCC was 0.035% and 0.17%, respectively (P = 0.292). The corresponding figures were corrected as 0.036% and 0.22% with SIR. PS-matching method produced 62 patients each in spontaneous and Nuc-treated HBsAg seroclearance groups. There was one patient with Nuc-treated HBsAg seroclearance but none with spontaneous HBsAg seroclearance developed HCC during a mean follow-up period of 80.8 and 80.9 months, respectively. There was no significant difference in the estimated annual incidence of HCC (0% vs. 0.24%, P = 0.606) between patients with spontaneous and Nuc-treated HBsAg seroclearance. None in both groups had the occurrence of death or hepatic complication.
Anti-HBs seroconversion after HBsAg seroclearance
The rate of anti-HBs seroconversion was 69.6% and 66.4% in spontaneous and Nuc-treated groups, respectively (P = 0.617) (Table 1). Among the PS-matched patients, the anti-HBs seroconversion was comparable (60.2% vs. 65.3%, P = 0.555) between two groups (Table 2). The 6-year cumulative incidence of anti-HBs seroconversion was 62.3% and 61.1% in spontaneous and Nuc-treated groups, respectively (P = 0.431) (Figure 1). The incidence reached plateau after 7-year follow-up.
image
Figure 1. The comparison of cumulative incidence of anti-HBs seroconversion between patients with spontaneous and Nuc-treated HBsAg seroclearance in propensity score-matched cohort. There is no significant difference in anti-HBs seroconversion between two groups (62.3% vs. 61.1%, P = 0.431). Nuc, nucleos(t)ide analogue.
Discussion
The interaction among HBV, hepatocytes and host immunity results in the distinct phases of the natural course in CHB infection.[1] HBsAg seroclearance usually represents sustained immune control of HBV and is the ultimate goal of antiviral therapy.[10, 28, 29] The results of the present study have shown that HBsAg seroclearance in Nuc-treated patients occurred at a significantly younger age (47.4 vs. 51.5 years), more frequently in those with cirrhosis but the clinical outcomes were similar as compared with spontaneous HBsAg seroclearance. Of note is that spontaneous HBsAg seroclearance usually occurs in patients more than 10 years after they have entered inactive phase.[2] In contrast, Nuc therapy is initiated in patients with active hepatitis at an age and the status of HBV infection not favouring spontaneous HBsAg seroclearance. This may imply that, despite HBsAg seroclearance rate is <1% per year, Nuc therapy may accelerate immune control and HBsAg seroclearance.
Both spontaneous and Nuc-treated HBsAg seroclearance in CHB patients indicate excellent clinical outcomes in previous studies.[3-8, 15, 17] As far as our knowledge, this is the first study with long-term follow-up period (mean >6 years) to explore whether there is difference in the clinical outcomes after HBsAg seroclearance between CHB patients with spontaneous and Nuc-treated HBsAg seroclearance. Previous studies have shown that old age (≥50 years) or cirrhosis at the time of HBsAg seroclearance, male gender and concurrent HCV or HDV infection are risk factors for HCC development after HBsAg seroclearance.[3, 7, 8] A total of five patients developed HCC at a mean follow-up period of 75.3 months after HBsAg seroclearance in the present study. All were male, three had pre-existing cirrhosis and all were older than 55 years at the time of HCC detection. As the direct effect of integration of HBV DNA into hepatocyte chromosomes is a possible aetiology of HCC development in CHB infection,[30] HCC can occur in the absence of cirrhosis after HBsAg seroclearance[7, 9] as observed in two patients in the present study. However, no histological evidence was available to confirm that these two patients had not developed cirrhosis prior to HBsAg seroclearance. Of the 110 Nuc-treated patients with HBsAg seroclearance, HCC developed in one (0.9%) patient with an estimated annual incidence of 0.14%. It is lower than 0.35% in a recent Korean study.[15] Younger age at the time of HBsAg seroclearance (mean 47.4 years) and predominant genotype B HBV infection in the present Taiwanese study might explain this discrepancy. The estimated annual incidence of HCC at 0.13% in patients with spontaneous HBsAg seroclearance in the present study is similar to the findings of previous studies.[3-9] Among each PS-matched cohort of 98 patients, the estimated annual incidence of HCC development was similar between spontaneous and Nuc-treated groups (P = 0.735) during a mean follow-up period of 76–97 months. The present study also showed comparable rates of death and variceal bleeding between two groups.
Since sustained remission of hepatitis during follow-up is significantly correlated with HBsAg seroclearance in the natural course[22] and that active necroinflammation and moderate to advanced histological fibrosis in liver are indications of antiviral therapy for CHB patients,[10, 28, 29] the patients with Nuc-treated HBsAg seroclearance were observed to have significantly higher proportion of cirrhosis than those with spontaneous HBsAg seroclearance in both overall population and PS-matched cohorts of the present study. Although there was no statistically significant difference, the estimated annual incidence of HCC was higher in Nuc-treated patients than those with spontaneous HBsAg seroclearance in overall population (0.17% vs. 0.035%) and in PS-matched cohort (0.24% vs. 0%) among the patients without cirrhosis at the time of HBsAg seroclearance. The patients with Nuc-treated HBsAg seroclearance might have higher possibility to develop HCC after longer time follow-up because of the pre-existing fibrosis before and during Nuc treatment.
The 6-year cumulative incidence of anti-HBs seroconversion in 110 Nuc-treated patients was 68.3%, which is similar to a rate of 68.9% in a recent Hong Kong study.[17] Among the PS-matched patients, there was no significant difference in the 6-year cumulative incidence of anti-HBs seroconversion between spontaneous and Nuc-treated groups (P = 0.431) (Figure 1).
Surveillance for HCC is cost-effective if the expected annual incidence exceeds 1.5% in cirrhotic patients, irrespective of its aetiology and 0.2% in patients with CHB infection.[26, 31] The recommended screening interval is 6 months and a shorter follow-up period (every 3–4 months) is suggested for cirrhotic patients in current guidelines.[26, 31] However, there is no consensus of the follow-up interval for those with HBsAg seroclearance.[3, 8, 9] Since cirrhosis, coinfection of HCV or HDV, male gender and old age (≥50 years) have been reported as independent risk factors of HCC development after HBsAg seroclearance,[3, 7, 8] periodic surveillance according to established clinical guidelines[26, 31] for detection of early HCC with ultrasound and AFP in subjects who cleared HBsAg with the above clinical characteristics is necessary. Considering cost-effectiveness, it seems reasonable to increase the interval of follow-up to 12 months after HBsAg seroclearance for subjects without the above risk factors.[3] Given that the clinical outcomes were comparable between spontaneous and Nuc-treated HBsAg seroclearance, perhaps similar surveillance policy is applicable to Nuc-treated patients with HBsAg seroclearance.
There are limitations in the present study. First, the number of patients in this retrospective study may not be large enough for the comparison in the clinical outcomes after HBsAg seroclearance between spontaneous and Nuc-treated patients. The small number of observed clinical events after HBsAg seroclearance, which was similar to the findings of previous studies,[3-9] also made the statistical power not strong enough to make a definite conclusion. Nevertheless, the mean follow-up duration of more than 6 years is long enough to observe the potential development of hepatic complications, even though much longer follow-up may be needed for assessing HCC development in the patients with Nuc-treated HBsAg seroclearance. Second, the discrepancy in age distribution and number size between patients with spontaneous and Nuc-treated HBsAg seroclearance may lead to bias in annual incidence and statistical comparison of clinical outcomes. The results have been corrected by SIR method and are consistent. Third, genotype was not assayed in the majority of patients in the present study due to unavailable stored serum or low/undetectable HBV DNA. However, the finding of predominant genotype B in our study population (71%) was comparable to the geographic distribution in Taiwan.[32] The genotype of the five patients with HCC development in the present study showed similar distribution (3 genotype B, 1 genotype C and 1 undetermined). Perhaps genotype is no more a factor for HCC development once patients lost HBsAg. Further longer follow-up studies may be needed to test the impact of HBV genotype on HCC development after HBsAg seroclearance.
In conclusion, the clinical outcomes between patients with spontaneous and Nuc-treated HBsAg seroclearance were similar. Higher rate of cirrhosis in Nuc-treated patients at the time of HBsAg seroclearance might reflect previously persistent hepatic necroinflammatory process. HCC can develop in a low rate during long-term follow-up after both spontaneous and Nuc-treated HBsAg seroclearance. HCC surveillance after HBsAg seroclearance using liver ultrasound and AFP is still mandatory.作者: StephenW 时间: 2016-4-15 15:52